Ulysses observations of field-perpendicular plasma flows in the Jovian magnetosphere: comparison of ExB velocity vectors derived from energetic ion and thermal electron data

The ExB velocity vectors which correspond to the field-perpendicular flow o f the low-energy plasma have been independently derived for the Ulysses fly by of Jupiter using energetic ion and thermal electron data. In this paper these measurements are compared. It is shown that reasonable quantitative a greement exists during the prenoon inbound pass, though with some significa nt differences. Both data sets indicate the presence of slow field-perpendi cular flows in the dayside outer magnetosphere (similar to 80-110 R-J), whi ch are directed azimuthally opposite to corotation with the planet and radi ally inwards, with magnitudes of similar to 100 km s(-1) in each component. Any variations in this flow during the outer magnetosphere traversal are n ot resolved within the similar to+/-100 km s(-1) uncertainties in the indiv idual 35 min-averaged data. Similar flows, but of somewhat smaller magnitud e, are also found in both data sets in the higher-latitude region of the in bound middle magnetosphere flanking the plasma sheet (similar to 45-70 R-J) . It is inferred that these field lines map equatorially into the outer mag netosphere at larger distances. Flows within the middle magnetosphere plasm a sheet in this region are on average in the sense of planetary rotation, b ut the average azimuthal velocity determined from the ATs data (similar to 20 km s(-1), essentially consistent with zero) is significantly smaller tha n that determined from the SWOOPS data (similar to 100 km s(-1)). A systema tic effect thus seems to be present within the current sheet, possibly asso ciated with additional (usually small) terms in the expression for the ener getic ion anisotropy which have not been taken into account in the analysis procedure. Analysis of the inbound data overall, however, indicates no con sistent velocity offset between these data sets to within a few tens of km s(-1), and a unit gradient between them within a factor of similar to 1.5. Due to the large uncertainties in individual similar to 35-min velocity val ues, however, and the possible current sheet effects mentioned above, the c ross-correlation coefficient between the data sets is low, with an overall value of 0.23 for the principal azimuthal component observed in magnetosphe ric regions over a 4-day interval on the inbound pass. The probability of t his degree of correlation appearing by chance, however, is only about one i n 500. On the outbound pass, we find that the velocity estimates determined from the two data sets do not agree, even qualitatively. We believe that t his is due to a complicated and anisotropic background in the electron data which we have been unable to fully remove. (C) 1999 Elsevier Science Ltd. All rights reserved.