A number of magnetopause crossings by the AMPTE/IRM spacecraft, which exhibited large shear in the magnetic field, were studied recently by Paschmann et al. (1986) in order to assess the presence or absence of reconnection processes at the dayside magnetopause. In the present article, eleven of these crossings are reanalyzed by use of methods described by Sonnerup et al. (1987). Although only eight of the cases lend themselves to this more detailed analysis procedure, the results derived when the analysis is successful are considerably more detailed and quantitative than those obtained previously. Four such cases are documented in detail in the paper. For these events it was possible to derive accurate vectors, n, normal to the magnetopause. It was also possible to place bounds on the values of average magnetic field and flow velocity components along n as well as on the average electric field tangential to the magnetopause and on the velocity of the magnetopause itself along n and therefore on the magnetopause thickness. It is particularly noteworthy that in two crossings (on September 4, 1984) it was possible to place lower bounds of 2.8 mV/m and 0.5 mV/m on the tangential (reconnection) electric field and corresponding lower bounds on normal flow and field. In a third crossing (September 8, 1984), located about 1 RE north of the subsolar point, evidence is deduced, indicating that the observed plasma acceleration along the magnetopause had a large temporal component, caused by a tangential gradient in the total pressure, in addition to the usual convective component caused by the Maxwell shear stresses. Finally, in the fourth crossing (October 19, 1984), large magnetic shear was present but, in agreement with a conclusion reached by Paschmann et al. (1986), the analysis shows that no evidence for reconnection was present and that the magnetopause had the properties of a tangential discontinuity. This case provides convincing evidence that the mere existence of a deHoffmann‐Teller frame of reference in which the flow is field aligned (an excellent fit of the data to such a frame could be performed in this case) does not guarantee that the magnetopause is a rotational discontinuity. The remaining events analyzed by use of the new procedure demonstrate that the methodology sometimes fails, or yields unconvincing results, even when there is considerable evidence, in the form of accelerated plasma flows, that some form of reconnection is in progress.
Read full abstract