We present high resolution images of the dust continuum optical
emission in the coma of comet Hale-Bopp on April 21-25 1997,
within a radius of 
of the photo-centre.
On processed images, coma detail is visible at characteristic
sizes and nuclear distances of
km or less. The dust pattern has been described
in terms of morphology and relative intensity, and is very complex.
It is shown that the filaments visible in the inner shell during
this period do not constitute the major part of the flux from the shell,
which is rather dominated by the extended diffuse emission.
From the cyclic appearance of features, a nucleus rotation period
of 
and a mean dust outflow terminal velocity
of 
has been determined. The north pole of the nucleus was
directed towards the Earth, as given by the sense of rotation of
the dust features.
It has been shown (e.g., by Kitamura 1986, 1987, 1990; Crifo et al. 1997a, 1997b, 1990) that the correlation between visible coma features and the location of the emitting nucleus source or sources is not trivial, especially not in the case of an active comet. Hydrodynamic modelling of gas and dust outflow reveal that an observed filament in the coma may be produced from the gas-dynamic interaction of emission activity from more than one source, and that higher-order daughter filaments are produced at larger radii by dust and gas interaction closer to the nucleus.
The fact that the outflow activity of a compositionally homogeneous
aspherical nucleus will produce shock fronts and filaments similar
to that of a compositionally inhomogeneous spherical nucleus,
indicates that coma structure may not give any intuitive information
on the location of the emitting source regions. Interpretation of
the three dimensional structure of the Hale-Bopp inner coma from
images such as those presented here in terms of nucleus activity
is further complicated by several factors, e.g., the integration
of the dust brightness flux along the line of sight, the probable
diurnal cycles of the sources, the probably complex shapes and
rotational states of the nucleus or nuclei, and the unknown dust
size and velocity distributions. We have an estimate on the size
of the dust grains from
Williams et al. (1997)
who found the optically important grains dominating the visual
scattering and near-infrared emission from the coma to have an
average radius of 
, based on infrared
spectrophotometric observations.
The value of the rotational period agrees well with the determinations of
earlier works. A longer period of observations would undoubtedly have
improved the error bars of the figure. A rotation period of
was determined from
from images obtained Jan. 12-Feb. 10 1997 (Lecacheux et al. 1997),
oscillating between 
and 
(Jorda et al. 1997).
A period of 
was obtained from images
taken at perihelion (Lisse et al. 1997)
and from mid-infrared images
(Sarmecanic et al. 1997a, 1997b),
from motion of dust arcs
(Farnham et al. 1998),
and 
by
Licandro
et al. (1998).
Long daylight observation
sequences by the European Hale-Bopp team in April 1997![[*]](/icons/foot_motif.gif){kind=icon}
indicates
a period of 
.
A possible longer period of emission activity of 20 days has
been reported (e.g., Sekanina 1996;
Jorda et al. 1997a;
Kidger 1998),
which has been attributed to precession of the nucleus. During the five day
time-frame of
the present observations, deviation from a pure spin rotational state
may, in addition to errors of measurement, contribute to the
error in the value of the obtained rotation period, which is based on
a "pure spin'' assumption. However, a very complex state of rotation of
the nucleus is contradicted by the work of
Licandro et al. (1998)
based on observations obtained during a time period similar to the length
of the suggested precession period.
The direction of the axis of rotation of the nucleus has been determined from Monte-Carlo image simulations and the application of different emission models (Sekanina & Boehnhardt 1997, 1998; Sekanina 1998), the results of which give somewhat contradictory results. It is possible, from results of gas-hydrodynamic modelling, that the rotational state of the nucleus may be very difficult to extract from imaging of coma features as filaments may be created by second-order interaction effects far from the source of emission, and thus not trace the movement of the nucleus.
The mean terminal expansion velocity of the dust shells determined here is
slightly higher than published values of 
from
CCD imaging in
Jan.-Feb. 1997 (Lecacheux et al. 1997),
and 
on Feb. 22, 1997
by Jorda et al.
From near-infrared observations, a velocity of the order of 
were determined by Mannucci & Tozzi (1997)
from observations on Feb. 3 and 10, 1997. The determination of
outflow velocity is uncertain partly due to the observing circumstances.
With a longer observing run and higher S/N data, a greater number of shells
would have been measurable over a longer time span, improving the estimate.
Further, this velocity estimate integrated over dust grain size and
line-of-sight effects is a mean value in a broad sense.
The complex structure of the filaments may be a result of variable activity of the active region(s) on the nucleus, or due to interaction of expanding outflowing dust volumes. We have limited ourselves to the morphological description of visible coma features, and hope that numerical simulation experiments and hydrodynamic modelling of the gas and dust interaction of the inner coma will benefit from detailed high resolution images such as those obtained by the SVST.
Acknowledgements
This work was made possible by a grant from the Anna and Allan Löfberg Foundation. Professor Göran Scharmer (Stockholm Observatory) is much thanked for providing observing time at the SVST. The observations would not have been possible without assistance from the SVST staff Göran Hosinsky and Rolf Kever. The comments of the anonymous referee were valuable in improving the paper.
Copyright The European Southern Observatory (ESO)