---> Last updated on 1 October 1997 <---
STIS is an acronym for the Space Telescope Imaging Spectrograph (STIS),
a "second generation" HST instrument installed during the most recent
servicing mission in February 1997. STIS has both imaging and
spectroscopic modes and operates from about 1200 Angstroms (=120
nanometers) to 10,000 Angstroms. Unlike the case for its two
predecessor spectrographs on HST, STIS spectra contain both spatial and
spectral information, which is very important in studying
spatially-extended objects like comets.
STIS first observed Hale-Bopp on 27 August 1997, which is when
the comet emerged from the HST solar exclusion zone. (HST could not
observe Hale-Bopp earlier in 1997 because the angle
between the sun and the comet was smaller than 50 degrees and
pointing HST that close to the sun could damage the telescope
and/or its instruments.) The next HST Hale-Bopp observations
are scheduled to occur during early November. Information on those
observations will also be posted at this web site.
You can access two times larger versions of all the figures appearing below by
clicking on them. We also provide access below to postscript versions of all
the full-size figures.
Fig. 1: The above image is a composite of two STIS CCD exposures taken
at 04:05 UT (2 sec integration) and 04:06 UT (20 sec integration) on 27
Aug 1997. The heliocentric and geocentric distances of the comet were
2.476 AU and 2.989 AU, respectively. The direction to the sun projected
onto this image is at 53.6 deg clockwise from the +x-axis (where the
+x-axis points to the right and the +y-axis points straight up), but
the solar phase angle (the sun-comet-earth angle) is 18 degrees, which
means that the direction to the sun is only 18 degrees from being
perpendicular to the plane of this image. A "long-pass" filter was used
that transmits all light longward of approximately 5500 angstroms (10
angstroms = 1 nanometer) and rejects light shortward of that
wavelength. The image is 28 arcsec on a side, which subtends 60,680 km
at the distance of the comet. Individual STIS CCD pixels are 0.0508
arcsec across, which projects to 110 km at the comet.
Click here to see the postscript version of
the above figure.
Fig. 2: The image in Figure 1 has been divided by a circularly
symmetric image that gives the best match to the azimuthally-averaged
surface brightness profile of the comet. This "ratio" image thus shows
the deviations of Hale-Bopp's dust production from spherically
symmetric, steady-state outflow. We see from the above image that
Hale-Bopp is still clearly showing the strong jet activity that started
becoming prominent towards late summer in 1996.
Click here
to see a detailed comparison of the September 1996 and August 1997
Hale-Bopp images.
Click here to see the postscript version of
the above figure.
Fig. 3: This figure is similar to Fig. 2, except that the
azimuthally-averaged image has been subtracted from the data.
This technique enhances the region near the nucleus, but otherwise
shows the same features as the ratio image given above.
Click here to see the postscript version of
the above figure.
Fig. 4: (a) shows the observed (boxes) azimuthally-averaged surface brightness
profile derived from the image. The solid curve is a power law surface
brightness profile with an index of -1, which is the expected value for
the case of spherically symmetric outflow of dust from a point source.
Also plotted (asterisks) is the point spread function (PSF) of the STIS CCD;
its magnitude at the peak pixel is the difference between the observed
peak pixel intensity and the extrapolated value expected from the coma.
Assuming that this PSF intensity represents light reflected from the nucleus,
and that the geometric albedo is 4%, the effective diameter of the nucleus
is approximately 60 km. However, this should be considered an upper limit
as strong temporal variability is clearly evident in the image from which
the profile is derived and some of the intensity at the peak pixel may be
due to dust produced during a newly-produced outburst. We continue to believe
that our estimate of the size based on the October 1995 HST image
(i.e., about 30-40 km) is the best that we can do using the HST data.
(b) shows the ratio of the observed surface brightness profile to the
model power law profile and illustrates where the observations most
strongly deviate from the steady-state case. The spike near the origin
is due to the presence of a "photometrically-resolved" nucleus and/or
a large outburst in dust production near the time of our observations.
Click here to see the postscript version of
the above figure.
Fig. 5: The above is a STIS spectral image taken using the CCD and the
G230MB grating and shows the detection of many lines in the OH(0,0)
band centered near 3090 angstroms. (Some weak lines in the OH[1,1] band
are also detected at the longer wavelengths.) The observations were
made through a long-slit of dimensions 2 arcsec (in the dispersion
direction, which is the horizontal dimension) by 50 arcsec (in the
spatial direction, which is the vertical dimension). The orientation
of this spectral image is exactly the same as for the images shown
above (i.e., celestial North is almost straight down).
From this spectrum we derive a PRELIMINARY value for the water
production rate of 3E29 molecules/sec, which is only slightly larger
than the values we derived from the HST observations last fall when the
comet was slightly farther from the sun. (On 23 Sep 1996 at r=2.97 AU
we estimated Q_water=2.1-2.6E29, while on 18 Oct 1996 at r=2.69 AU we
obtained 2.7E29). In addition, we now have detailed information on the
spatial distribution of the OH and find (again this is PRELIMINARY)
that it is similar to that predicted by the standard vectorial model.
The intensities of the individual lines in the OH band appear to
deviate somewhat from that expected in fluorescence equilibrium, and we
are currently investigating this issue. The two dark horizontal streaks
at approximately 12 arcsec above and below the continuum emission are
due to support structures that stabilize the slits. For the raw data,
the dispersion is 0.15 angstroms/pixel and the plate scale is 0.0508
arcsec/pixel, but we have re-binned the data by a factor of four in
order to improve the signal-to-noise ratio (i.e., the effective scales
in the above spectrum are 0.6 angstroms/pixel and 0.20 arcsec/pixel).
The total exposure time for this image was 28 mins, centered at 02:14
UT on 27 Aug 1997.
Click here to see the postscript version of
the above figure.
Fig. 6: Here we have extracted a 1-dim spectrum from the G230MB
spectral image by summing over 2 arcsec in the spatial dimension,
centered on the continuum (i.e., +/- 1 arcsec about the continuum; the
effective aperture size is 2 arcsec x 2 arcsec). A scaled solar
spectrum that best matches the cometary continuum is overplotted in
green. The prominent features are the strongest lines in the OH(0,0)
band. The quoted absolute fluxes are PRELIMINARY.
Click here to see the
postscript version of the above figure.
Fig. 7: This net spectrum results from subtracting the solar spectrum
from the cometary spectrum (i.e., differencing the two curves in figure
6). The prominent features are the strongest lines in the OH(0,0) band.
Weak emission in the OH(1,1) band can also be seen between 3135 A and 3160 A.
The quoted absolute fluxes are PRELIMINARY.
Click here to see the
postscript version of the above figure.
Fig. 8: The above is a STIS spectral image taken using the CCD and the
G230LB grating. The edge of the strong OH(0,0) band is seen in the far
right columns. We have TENTATIVELY identified the CS(0,0) band near
2580 angstroms and the OH(1,0) band near 2826 angstroms. For the raw
data, the dispersion is 1.3 angstroms/pixel and the plate scale is
0.0508 arcsec/pixel, but we have re-binned the data by a factor of four
in order to improve the signal-to-noise ratio (i.e., the effective
scales in the above spectrum are 5.2 angstroms/pixel and 0.20
arcsec/pixel). Although this grating actually covers a spectral range
from 1685 angstroms to 3065 angstroms, detectable signal from Hale-Bopp
is only present longwards of about 2200 angstroms. The total exposure
time for this image was 30 mins, centered at 03:42 UT on 27 Aug 1997.
Click here to see the postscript version of
the above figure.
Fig. 9: Here we have extracted a 1-dim spectrum from the G230LB
spectral image by summing over 2 arcsec in the spatial dimension,
centered on the continuum (i.e., +/- 1 arcsec about the continuum; the
effective aperture size is 0.5 arcsec x 2 arcsec). The reddened solar
spectrum that best matches the cometary continuum is overplotted in
green. The quoted absolute fluxes are PRELIMINARY.
Click here to see the postscript version of
the above figure.
Fig. 10: This net spectrum results from subtracting the solar spectrum
from the cometary spectrum (i.e., differencing the two curves in figure
9). We feel that we have detected the CS(0,0) band near 2576 A and the
OH(1,0) band near 2820 A, but only marginally. The feature near 2713 A
appears to be an artifact. The quoted absolute fluxes are PRELIMINARY.
Click here to see the
postscript version of the above figure.
Paul Feldman, Johns Hopkins University
Mike A'Hearn, University of Maryland
Claude Arpigny, Universite de Liege (Belgium)
Jack Brandt, University of Colorado
Alan Stern, Southwest Research Institute (Boulder Extension Office)
Melissa McGrath, Space Telescope Science Institute (STScI Contact Scientist)
Andy Lubenow, Space Telescope Science Institute (STScI Program Coordinator)
*I would also like to acknowledge the generous ephemeris support provided
by Don Yeomans and the JPL Solar System Dynamics Group and by Brian
Marsden and the MPC.
You can e-mail comments on the above to weaver@pha.jhu.edu .