ION INJECTION AND FERMI ACCELERATION AT EARTHS BOW SHOCK - THE 1984 SEPTEMBER 12 EVENT REVISITED

The injection and acceleration of thermal solar wind ions at the quasi-parallel bow shock is investigated. Observations made by the Active Magnetospheric Particle Tracer Explorers/Ion Release Module (AMPTE/IRM) on 1984 September 12 near the nose of the bow shock have recently been reported by Ellison, Mobius, & Paschmann and were modeled by these authors by a Monte Carlo simulation which includes diffusive shock acceleration. It was concluded that the ions are injected into the acceleration process by thermal leakage from the downstream plasma. We have modeled this event by large-scale self-consistent one-dimensional hybrid simulations. Two different simulation runs were performed. In run A, the simulation macroions consisted only of protons. In order to obtain reasonable statistics of the accelerated ions each backstreaming ion is split into a large number (30) of new ions which were subsequently followed in space and time. We find that the spectral observations upstream of Earth's bow shock are in good agreement with the result of the hybrid simulation. It should be emphasized that there is no free parameter and there are no additional assumptions in these simulations once the upstream solar wind conditions are specified and the upstream free escape boundary is fixed. Protons up to approximately 50 keV are accelerated at the shock out of the incident thermal solar wind distribution by a more or less coherent mechanism. Ions above approximately 50 keV have been scattered in the upstream region, returned to the shock, and interacted with the shock at least a second time. Run B was performed in a smaller simulation system; however, alpha particles were included self-consistently in such a way that sufficient simulation particles are present to give reasonable statistics. The upstream spectra of both protons and alpha particles can be modeled by the hybrid model when it is assumed that the solar wind contains during this time period approximately 7% alpha particles and the alpha particles have the same temperature per nucleon as the protons. The agreement between the observations and the simulations is less satisfactory in the downstream; this is attributed to the one-dimensionality of the simulations. The alpha particle contribution is considerably higher than that needed by Ellison, Mobius, & Paschman. This play provide a critical test for the two models.