We have shown that the plasma QTN measured by Ulysses plummets in the Bernstein
forbidden bands
.
As a consistency check, we have verified that the onboard relaxation sounder
spectra exhibits resonances at about the same frequencies as determined nearly
simultaneously from the QTN spectra. To the best of our knowledge, these
forbidden bands have never been detected before, presumably because of the lack
of sensitivity and frequency resolution of earlier instruments. In particular,
such a detection could not be performed with the Voyager 1 spacecraft radio
data from which [Birmingham et al., 1981] derived the electron density in
the Io plasma torus. Their method was based on detection of the upper hybrid
resonance emissions which they expected to produce the strongest peaks in the
power spectra (other methods used to determine the Jupiter's electron density
from the plasma wave observations are summarized by [Gurnett et al., 1981]).
Using this method, [Hoang et al., 1993] have given the
electron density along the Ulysses trajectory inside 
, identifying
in the high band ( 
50 to 1000 kHz through only 12 channels) of
the URAP radio receiver in the region where 
. Note that a
particular relevance of all these electron density measurements is that they are
unaffected by spacecraft charging or sheath effects. A further advantage of the
method introduced in the present paper is that the detection of stop bands
allows the location of the 
without ambiguity. The method of deducing
from the strongest peak may be somewhat precarious for 
when the resonance peaks at a higher level, which may happen for
some antenna geometries [see, e.g., [Christiansen et al., 1978], Figure 3].
In addition, depends on many plasma parameters,
whereas the 
are only functions of the main (cold) electron
population density 
, and so their determination allows us to deduce .
The uncertainty in is about 16% and mainly due to the uncertainty in
the Doppler shift produced by the plasma corotation.
Hence this method of QTN
spectroscopy allows us to routinely measure in situ the core plasma
density every 
min (or 
) along that part of Ulysses
trajectory between 9 and 
. The results are shown in
figure 5.
The measurement gaps (about 20% of the spectra) are due to pollution by Jovian
radio emissions (near 2030 UT) and presumably to high densities bringing
the lowest outside our spectral range (near 1830 UT). Let
us finally recall that this determination of is based on the description
of the electron distribution as a superposition of two Maxwellians. There are,
however, some indications that the distribution there might be, instead,
kappa-like [Meyer-Vernet, Moncuquet and Hoang, 1995]. In such a case, the
total density can still be estimated from the present analysis, since the first
forbidden band can serve to localize the intraharmonic band containing
. A more accurate measurement of the electron density would require
calculating the with the actual electron distribution function; this
deserves further investigation.
Figure 5: Core electron density deduced from
Bernstein wave forbidden bands along Ulysses trajectory in the outskirts of
the Io plasma torus. The bottom axis shows the Jovicentric distance, and the
top axis shows the distance from centrifugal equator (determined from the
Goddard Space Flight Center 
magnetic field model).
Acknowledgments
The URAP experiment is a joint project of NASA GSFC, Observatoire de Paris,
CETP, and the University of Minnesota. The French contribution is supported by
the Centre National d'Études Spatiales and the Centre National de la
Recherche Scientifique. Support for the magnetic field investigation at
Imperial College is provided by the U.K. Particle Physics and Astronomy
Research Council. We sincerely thank J.-L. Steinberg and F. Bagenal for a
careful reading and helpful comments on the manuscript.
The Editor thanks M. Ashour-Abdalla and another referee for their
assistance in evaluating this paper.