Magnetic turbulence at the magnetopause, a key problem for understanding the solar wind/magnetosphere exchanges

According to ideal MHD, the magnetopause boundary should split the terrestrial environment in two disconnected domains: outside, the solar wind (including its shocked part, the magnetosheath), and inside, the magnetosphere. This view is at variance with the experimental data, which show that the magnetopause is not tight and that a net transfer of matter exists from the solar wind to the magnetosphere; it implies that the frozen-in condition must break down on the magnetopause, either over the whole boundary or at some points. In the absence of ordinary collisions, only short scale phenomena (temporal and/or spatial) can be invoked to explain this breakdown, and the best candidates in this respect appear to be the ULF magnetic fluctuations which show very strong amplitudes in the vicinity of the magnetopause boundary. It has been shown that these fluctuations are likely to originate in the magnetosheath, probably downstream of the quasi-parallel shock region, and that they can get amplified by a propagation effect when crossing the magnetopause. When studying the propagation across the magnetopause boundary, several effects are to be taken into account simultaneously to get reliable results: the magnetopause density gradient, the temperature effects, and the magnetic field rotation can be introduced while remaining in the framework of ideal MHD. In these conditions, the magnetopause amplification has been interpreted in term of Alfven and slow resonances occurring in the layer. When, in addition, one takes the ion inertia effects into account, by the way of the Hall-MHD equations, the result appears drastically different: no resonance occurs, but a strong Alfven wave can be trapped in the boundary between the point where it is converted from the incident wave and the point where it stops propagating back, i.e., the point where k(parallel to)=0, which can exist thanks to the magnetic field rotation. This effect can bring about a new interpretation to the magnetopause transfers, since the Hall effect can allow reconnection near this particular point. The plasma transfer through the magnetopause could then be interpreted in terms of a reconnection mechanism directly driven by the magnetosheath turbulence, which is permanent, rather than due to any local instability of the boundary, for instance of the tearing type, which should be subject to an instability threshold and thus, as far as it exists, more sporadic.