Recent results on reconnection onset in thin current sheets have shown that the underlying physics depends upon the nonlinear interaction of micro-instabilities, and that these can be very challenging for standard, explicit simulations to do even in two dimensions. Implicit simulations using the implicit moment method showed that the lower hybrid drift instability can accelerate tearing mode growth such that fast reconnection occurs. These results in three dimensions, which were obtained on a commodity workstation, depend nonlinearly on physics that can be studied in two dimensions, which puts them just within reach of confirmation by explicit simulations. These were performed on 128 nodes of a massively parallel computer, and confirm the implicit results . To connect these results with the real magnetotail requires the addition of a perpendicular field, which means open boundary conditions for the simulation since B normal will be nonzero. Since the electrons are so mobile, inaccurate boundary conditions can influence the solution at the neutral sheet unless the boundaries are sufficiently distant. Even with the implicit moment method, the cost of a domain that provides isolation and resolves near-electron scales is formidable. It is argued that a property of the implicit moment method offers an opportunity for a real advance in the capability to do problems of this type. As the grid spacing increases with the ratio of time step to grid spacing held constant, solutions remain stable and asymptote to a two-fluid, quasi-neutral system. With collisions, the implicit moment equations become the Hall-magntohydrodynamic equations. This two-fluid system, in which the electron current is computed from an Ohms law, are compatible in dependent variables and characteristics with the kinetic equations. It is proposed first to implement the Hall-MHD limit of the implicit moment equations, and to verify that it will give correct results for standard problems, such as the GEM reconnection challenge. Second, interface conditions between kinetic and Hall-MHD regions will be developed. These require continuity in the field solutions, flux matching conditions for the fluid moments, and consistent boundary conditions for kinetic particles. These will be tested for uniform flows, and for a resolved kinetic calculation of the lower hybrid drift instability enclosed within a stretched Hall-MHD grid. Third, current sheets threaded by open field lines will be modeled in two and three dimensions with special attention to the sensitivity to MHD domain outer boundary conditions, and the role of parallel currents.
