Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stsci.edu/ftp/meetings/jupiter-impact/poster_abstracts.ps Äàòà èçìåíåíèÿ: Wed Apr 26 18:31:54 1995 Äàòà èíäåêñèðîâàíèÿ: Sat Dec 22 19:10:22 2007 Êîäèðîâêà:

Evolution of the SL9 Impact Sites and Numerical Experiments
with a Two­Layer Rigid­Lid Model
Tadashi Asada
Kyushu International University
Isao Miyazaki
Oriental Astronomical Association
Tokuhide Akabane
Hida Observatory
Ground­based imaging of impact sites of Comet Shoemaker­Levy 9 on Jupiter was carried
out in visible and near infrared wavelength. Many impact sites were found to be elongated
in east­west direction due to the shear of the zonal #ow.
Numerical experiments with two­layer rigid­lid model indicate that an eddy evoluves to be
elongated in east­west direction by the horizontal shear of the zonal #ow in the case of
small Rossby deformation radius, whereas it survives keeping its circular shape in the large
deformation radius case. The evolution of impact sites seems to resemble the behavior of the
eddy of small defomation radius.
1

IR­Images of Calar Alto: Investigation of the temporal development of the SL9 impact
sites on Jupiter
Johannes Babion (University Observatory, Munich)
Hermann Boehnhardt (University Observatory, Munich)
Fionn Murtagh (ST­ECF ESO, Garching)
Jean L. Starck (CEA, DSM/DAPNIA, CE­SACLAY, 91191 Gif­sur­Yvette Cedex)
The SL9 impacts at Jupiter were imaged from Calar Alto (Spain) with the 3.5 m telescope
+ MAGIC Camera in the near infrared and with the 1.2 m telescope + CCD in the visual
wavelength range. After the basic data reduction suitable deconvolution techniques (Babion
et al, European SL9/Jupiter Workshop Proceedings, 1995) are applied to the images to
sharpen the surface structures of the impact regions on the planet. Using our time series
of multi­wavelength observations the temporal development of the SL9­impact is assessed.
This still ongoing analysis adresses the rapid expansion during the initial phase after impact
and the long­term diffusion of the ejecta structures on Jupiter. In this paper a report on the
present status of the project is given and preliminary results are described.
2

Dynamical Evolution of Comet P/Shoemaker±Levy 9
M.E. Bailey \Lambda , V.V. Emel'yanenko \Lambda & J.V. Scotti \Lambday
\Lambda School of Computing and Mathematical Sciences, Liverpool John Moores University,
Byrom Street, Liverpool L3 3AF, England.
y Lunar and Planetary Laboratory, University of Arizona, Tucson, Arizona 85721, USA
We report on both the analytical and numerical investigation of the dynamical evolution of
P/Shoemaker­Levy 9. Secular perturbations on objects in temporary­satellite­capture (TSC)
orbits about Jupiter cause frequent evolution on to orbits that allow tidal disruption or
collision with the planet. The effect is analogous to the Sun­grazing phenomenon amongst
high­inclination Halley­type orbits about the Sun, and it produces a much higher Jupiter­
collision rate than that expected from a random encounter model. The order of magnitude
of the collision rate is the probability (¸ 10 \Gamma1 ) for a long­lived TSC object to evolve into a
Jupiter­grazing or Jupiter­colliding end­state by secular perturbations, multiplied by the rate
of production of such TSC events per observable Jupiter­family comet per year (¸2 \Theta 10 \Gamma4 ).
Secular theory and direct numerical integrations show that the fragments of P/Shoemaker­
Levy 9 passed through a previous extremely close encounter with Jupiter in 1970, suggesting
that the progenitor must also have made a close approach to Jupiter around this time. We
discuss the similarity between this dynamical evolution and that found for variational orbits
based on P/Gehrels 3 and P/Helin­Roman­Crocket. Remarkably, P/Gehrels 3 was also within
the Roche radius of Jupiter at the same time as P/Shoemaker­Levy9 in 1970. Backward
integrations of test particles with initial orbits similar to those of the P/Shoemaker­Levy9
fragments are used to investigate the distribution of possible source orbits for the progenitor
of P/Shoemaker­Levy 9.
3

Far­ultraviolet emissions from the impact sites of comet P/Shoemaker­Levy 9
with Jupiter
G.E. Ballester(1), W.M. Harris(1), G.R. Gladstone(2), J.T. Clarke(1),
R. Prange(3), C. Emerich(3), A. Talavera(4), S.A. Budzien(5),
P.D. Feldman(6), M.R. Combi(1), T.A. Livengood(7), M.B. Vincent(1),
M.A. McGrath(8), D.T. Hall(6), D.F. Strobel(6), J.M. Ajello(9),
L. ben Jaffel(10), D. Rego(1), G.F. Fireman(11), K.L. Jessup(1),
L. Woodney(12), S. Miller(13), and X. Liu(9)
(1) U. of Michigan, (2) Southwest Res. Inst.,
(3) Inst. d'Astrophysique Spatiale, France, (4) IUE­VILSPA, Spain
(5) Naval Res. Lab., (6) Johns Hopkins U., (7) Goddard Space Flight Center
(8) STScI, (9) Jet Propulsion Lab.
(10) Inst. d'Astrophysique de Paris, France, (11) Computer Sciences Corp.
(12) U. of Maryland, (13) U. College London, UK
Observations with the International Ultraviolet Explorer (IUE) during the collisions of frag­
ments of comet P/Shoemaker­Levy 9 with Jupiter show far­ultraviolet emissions from the
impact sites within a 20 minute period centered around impact time. Positive detections
of H2 Lyman­ and Werner­band (1230­1620 A), H­Ly a, C 1657 A, and possibly Al+ 1671
A emissions were made for the K and S fragment impacts. Energies of at least 10(21) ergs
were emitted at these wavelengths, corresponding to brightnesses of at least a few hundred
kilo­Rayleighs. No thermal far­UV continuum was detected. The H2 spectrum is consistent
with electron collisional excitation if signi®cant CH4 absorption is included. Such emis­
sions could result from magnetospheric and/or ionospheric plasma processes generated by
the energetic impacts, and possibly related to the far­UV emissions imaged 40­50 minutes
after the K impact with HST. In this emission process, the hydrogen could be atmospheric
while the C and Al+ would have to be of cometary origin, and the CH4 absorption could
be easily explained by a stratospheric column in the line­of­sight. Another type of emission
process involves resonant and #uorescent scattering of continuum thermal radiation by upper
atmospheric hydrogen and ablated comet material. The hot thermal far­UV radiation could
have been produced by either the later stage of the fragment entry, when reaching suf®ciently
high pressure levels where the shock may have reached very high temperatures, and/or also
by the early ®reball which may have reached temperatures above 3500 K when at low alti­
tudes and behind the limb (cooling to below 1800 K when directly observable from Earth).
Only small quantities of cometary C and the possible Al+ would be needed for this emission
process. The plausibility of this process depends on the shape of the H2 spectrum, which has
not been modeled. Scattering of sunlight by the rising plume cannot explain the emissions,
except possibly for the Al+ emission, and direct collisional excitation in the earliest entry
stage or non­thermal emission by the early non­equilibrium entry shock could in principle
have contributed to the emissions.
4

2¯m spectro­photometry of Jovian aerosol clouds ± scattering opacities, vertical
distributions and wind velocities
D. Ban®eld, P. Gierasch, S. Squyres, & P. Nicholson (Cornell)
B. Conrath (Goddard), and K. Matthews (Caltech)
Spectro­photometric observations of Jupiter at 9 wavelengths in the 2.0 ± 2.35 micron band
were taken on the 5m Hale telescope at Palomar during the 2 months following the impact of
comet P/Shoemaker­Levy 9. From these, spectra of the discernable spots were measured as
they evolved through this time. Spectra of 4 regions (South Polar Region, North Equatorial
Zone, South Equatorial Zone and Great Red Spot) unaffected by the comet impacts were
also measured. A technique with which near­infrared re#ection spectra can be inverted to
yield vertical pro®les of scatterer density is presented. For the wavelengths and bandwidths
measured in this study, the sensitivity of the inversions extends from near 1 bar up to about
15mbar. This technique was used to estimate the height and total scattering opacity of the
measured comet impact clouds, with the result that all comet impact clouds were found to
lie near or above the vertical limit of our sensitivity, 15mbar. The lower limit of the clouds
is around 50mbar. The total scattering opacity of all of the clouds is seen to decrease by a
factor near 1.7 over the 35 days spanned by the observations. This is consistent with a change
in the mean particle size, with a general fall­out of some of the cloud material, and with a
horizontal dispersion of clouds. West et al. (1995) suggest that the mean particle radius
does increase over this time. We have not measured the horizontal dispersion of the clouds
over this time period, and thus cannot yet separate out the three possible effects to examine
particle fall­out rates. The speeds of stratospheric winds experienced by these clouds were
also measured by two different techniques. The ®rst assumed an a priori latitudinal pro®le
for the stratospheric winds (the tropospheric winds of Limaye (1986)) which was then scaled
down. The scaling factors were chosen to deform early images of the clouds so that they
best matched later images of the same clouds. The second technique imposed no a priori
assumptions on the wind ®eld, but rather measured it via a technique similar to that used
in Limaye(1986). Early and later images were correlated with each other as a function of
longitudinal shift for each latitude, yielding zonal velocities given by the shift of maximum
correlation. Both techniques indicate wind speeds of order ­5±15m/s (¸ 1=2 those measured
by Voyager in this region). The spectra of the unaffected regions of Jupiter are a very useful
by­product of this data set. They can be used to verify calibrations, as well as to quantify the
vertical distribution of aerosols in undisturbed regions. For the South Polar Region, we ®nd a
distribution of aerosols (total scattering optical depth ¸ 0:05) peaking at about 25mbar, and
extending down to near 100mbar. We are not sensitive to its top limit. The North and South
Equatorial Zones show what appears to be a diffuse aerosol cloud (total scattering optical
depth ¸ 2 \Theta 10 \Gamma3 ) near or above the vertical limit of our sensitivity, overlying a perhaps
higher tropopausal cloud deck at around 450mbar. The South Equatorial Zone shows more
variability than the North. The Great Red Spot shows only a tropopausal cloud which is
again slightly elevated compared to the rest of the planet, peaking at about 400mbar. Our
re#ectance spectra inversion technique can also be applied to other regions of the planet to
investigate the distribution of aerosols there as well.
5

On the Orbital Evolution and Origin of Comet Shoemaker­Levy 9
L.A.M. Benner and William B. McKinnon
Dept. Earth and Planetary Sci. and McDonnell Center for the Space Sci.
Washington University
Saint Louis, MO 63130
(e­mail: mckinnon@wunder.wustl.edu)
Based on our dynamical study (submitted for publication), we ®nd that Shoemaker­Levy 9
(SL­9) orbited Jupiter for most of this century, and possibly longer, by far the longest known
temporary capture by Jupiter. Secular solar perturbations caused SL­9's inclined and distant
jovicentric orbit to alternate, with a period of ¸25 yr, between intervals of nearly polar
inclination (i ¸ 90 ffi ) and low eccentricity and intervals of lower inclination (either prograde
or retrograde) and nearly parabolic eccentricity (e ú 1.0). This latter phase brought the comet
close enough for disruption on 8 July 1992 and collision in July 1994. Immediately prior to
the 1992 perijove, the comet passed through Jupiter's tenuous inner ring halo at 1:62 \Sigma 0:01
jovian radii, but the population of macroscopic bodies in this portion of the ring is probably
insuf®cient for collisions to have weakened the nucleus or disrupted the nucleus. The comet's
trajectory was among the most chaotic of any known solar system body, with an effective
Lyapunov time in jovian orbit of ¸10 years. Because of the shortness of the Lyapunov time
compared to the probable length of temporary capture about Jupiter, SL­9's heliocentric orbit
prior to capture can only be discussed in a statistical sense.
The ensemble of possible pre­capture solutions describe low e and i orbits just exterior and
interior to that of Jupiter. These orbits are indistinguishable from those of "quasi­Hilda­
type" Jupiter­family comets, a group known to strongly interact with Jupiter and experience
extended temporary satellite captures. Based on their Tisserand parameters, the orbits are not
consistent with a recently escaped Trojan asteroid, but an origin as such is not impossible.
6

Mid±IR High±resolution Spectroscopy of the SL9 Impact Sites: Temperature and
HCN Retrievals
B. B# ezard (Paris­Meudon Obs.), C.A. Grif®th (N. Arizona U.), D. Kelly (U. Texas), J. Lacy
(U. Texas), T. Greathouse (N. Arizona U.), and G. Orton (NASA/JPL)
We present observations of the SL9 comet's crash that were taken at the NASA Infrared
Telescope Facility (Mauna Kea, Hawaii), with the U. Texas mid±infrared echelle spectrometer
IRSHELL. Various impact sites were imaged at a spatial resolution of 1.5 arcsec and a
resolving power of about 15000. Several spectral intervals in the 10±¯m window were
observed. CH 4 lines around 1234 cm \Gamma1 and a C 2 H 2 line near 743 cm \Gamma1 were observed on
July 20, 30 and 31 to monitor stratospheric temperatures and constrain the analysis of the
observed NH 3 emission lines (see poster by Grif®th et al.). CH 4 line spectra recorded 23
hrs after impact K and 11 hrs after impact L show enhanced emission over these sites. This
enhancement is very pronounced for the 1233.455±cm \Gamma1 line but negligible for the ¸100
times weaker lines at 1233.147 and 1234.226 cm \Gamma1 , implying that the stratospheric warming
was con®ned to levels above ¸0.2 mbar. We derive temperature enhancements around 45 K
for the L impact and 25 K for the K impact at 1 ¯bar at the time of the observations. Mapping
of the C 2 H 2 line at 743.265 cm \Gamma1 indicates a local warming of ¸10 K in the 10±¯bar region
over the E site, 2.6 days after impact. HCN emission was detected at 744.458 cm \Gamma1 over
all impact sites observed. The retrieved HCN column density is somewhat sensitive to
assumptions in the gas vertical distribution and in the temperature pro®le. A preliminary
analysis indicates a column density of about 10 16 molec cm \Gamma2 at the center of the E site
2.6 hrs after impact, 3 times larger than inferred from millimeter observations of impact G
(Marten et al. 1995).
7

Detection of Water in the "Splash" of Fragments G and K of Comet Shoemaker­Levy 9
G. L. Bjoraker
NASA/GSFC
T. Herter, G. Gull, S. Stolovy, and B. Pirger
Cornell University
We observed Jupiter on 17­19 July 1994 using the Kuiper Airborne Observatory (KAO)
deployed out of Melbourne, Australia. The KAO Echelle Grating Spectrometer uses a Si:As
BIB array with 128 spectral elements at a resolving power of 9000 and 10 spatial elements on
Jupiter. We detected H2O emission lines at 7.7, 22.6, and 23.9 microns. We have obtained
temperatures at the peak of the splash phase for the G and K fragments, for the 3 microbar
pressure level. The H2O lines at 7.7 microns were visible for 10 minutes; those at 23.9
microns were visible for about an hour in G. We have also obtained column abundances for
H2O for both G and K using 7.7 micron data. Our estimate should be considered a lower
limit to H2O as the 7.7 micron data are sensitive only to the portion of water at the highest
temperatures. The G and K fragments exhibited remarkably similar spectra at the peak of
the splash at 7.7 microns. The column abundance of "hot" H2O at the peak of the splash was
roughly comparable to that of CH4. We interpret the CH4 emission as due to pre­existing
jovian CH4 heated by infalling ejecta. We believe that the oxygen in the observed H2O is
derived from cometary material, although we cannot distinguish between H2O that may have
been formed through shock chemistry from H2O that may have survived the initial explosion.
A combination of Galileo/NIMS, KAO/HIFOGS, and KAO/KEGS data favor a scenario in
which the fragments did not reach the water cloud in order to excavate jovian H2O.
8

EFFELSBERG OBSERVATIONS OF JUPITER'S RADIO BRIGHTNESS
VARIATIONS INDUCED BY COMET SHOEMAKER­LEVY 9
M.K. Bird & O. Funke
Radioastronomisches Institut
Univiversit# at Bonn
Auf dem H# ugel 71, 53121 Bonn, Germany
J. Neidh# ofer
Max­Planck­Institut f #
ur Radioastronomie
Auf dem H# ugel 69, 53121 Bonn, Germany)
I. de Pater
Dept. of Astronomy
University of California
Berkeley, CA 91109, U.S.A.
As part of a worldwide campaign of Jovian radio observations associated with the impact of
comet SL9, measurements of Jupiter's #ux density and polarization were recorded for three
wavelengths (– = 2.8, 6, 11 cm) at the Effelsberg 100­m Radio Telescope of the Max­Planck­
Institut f#ur Radioastronomie. The nonthermal #ux densities, attributed to gyrosynchrotron
radiation from energetic electrons in the Jovian radiation belts, are normalized to 4.04 AU
and corrected for the ®nite size of the emitting region with respect to the beamwidth, an
effect which can be quite important at the higher frequencies. Two parameter ®ts were made
for each observation day to the Jovian beaming curve, the radio #ux density as a function of
Jupier's rotational phase. The value of the constant term, the #ux density S 0 , increased from
4.4 Jy to 5.7 Jy (27%) at – = 11 cm and from 2.4 Jy to 3.9 Jy (62%) at – = 6 cm during the
impact week from July 16 to July 22. No changes in the radio brightness of 25.6 \Sigma 0.8 Jy
could be detected at – = 2.8 cm. The brightness level did not rise uniformly, but was pushed
up preferentially at speci®c System III longitudes. Most recent observations in February
1995, more than six months after the the SL9 collision, indicate that Jupiter's nonthermal
radiation has still not returned to its pre­impact level.
9

Europe and Io photometry during SL9 fragments impacts
C. Blanco 1 , G. Leto 1;2 , D. Riccioli 1
1 Astronomy Institute of Catania University
Viale A.Doria 6, I 95125 Catania, Italy
2 Institute of Radioastronomy, VLBI Station
C.P. 169, I­96017 Noto (SR), Italy
We report on the photometric monitoring of Europa and Io during the impacts of the comet
P/Shoemaker­Levy 9 fragments on Jupiter. BV photoelectric observations were collected
at the Serra La Nave (SLN) Station of Catania Astrophysical Observatory. High speed
photometry at 0.1 sec and 1 sec integration time was performed by using the 91­cm Cassegrain
telescope equipped with a photon counting photometer provided by a cooled photomultiplier.
Constraints due to SLN geographical position have limited our observations to 3 impact events
(A, H, Q family). For A and Q events we have recorded, for Europa and Io respectively,
light brightning exceeding two sigma. The times of the events, their duration and the energy
budget are given and discussed in the light of other observer results.
10

Radiation study of two very bright terrestrial bolides
J. Borovi# cka and P. Spurn# y
Astronomical Institute, 251 65 Ond# rejov Observatory, Czech Republic
For the comparison with the entry phase of the SL­9 collision with Jupiter, light curves
and spectra of two of the brightest photographed bolides were studied. The #
Sumava bolide
(EN 041274) reached \Gamma21 absolute (i.e. 100 km distance) magnitude. This fragile cometary
body of the initial mass of about 5000 kg exhibited many #ares and disappeared at the height
of 59 km above surface. The light curve can be explained assuming that the body experienced
at least ®ve subsequent sudden breakups, losing from 25% to 99% of its instantaneous mass
in each breakup.
The Bene#sov bolide (EN 070591, maximal magnitude \Gamma20) was a stony body of the initial
mass of about 13 000 kg which radiated down to 17 km of height and exhibited bright #are
at 24 km. Several fragmentation events occurred below 42 km of height and are directly
visible on the photographs. In both cases the spectra are composed of emission lines of the
vaporized material. In addition, molecular bands are strong in Bene#sov.
11

Near­Infrared Spectroscopy of the Shoemaker­Levy 9 Impact Sites with UKIRT: CO
Emission from the L Site
T.Y. Brooke, G.S. Orton, D. Crisp, A.J. Friedson(JPL/Caltech), G. Bjoraker(Goddard)
Spectra of impact sites in select regions in the 3­5 ¯m range were obtained with the CGS4
spectrometer at the United Kingdom Infrared Telescope atop Mauna Kea, Hawaii on July 19,
20, 26, 28 UT 1994, generally long after actual impact. These are being used to determine
the depth of impact effects in Jupiter's atmosphere.
CO emission in the 4.7 micron fundamental vibrational band was detected at the L site on
Jul 20 from 2:20­3:30 UT, over 4 hours after impact. Lines observed in emission were P4­P6
and R11­R13. Phosphine and CH 3 D were not seen in emission. Modelling is in progress to
determine CO abundances at the L site.
12

Limits on Polar Ionospheric Changes during the SL9 Impacts
Michael E. Brown
Lunar and Planetary Laboratory
Proposed explanations for many of the auroral and magnetospheric effects thought to be
observed as a consequence of the SL9 impacts rely on a changes in polar ionospheric
conductivity to transmit the effects through the magnetosphere. Observations of the Io
plasma torus rotation velocity, which is controlled by this ionospheric conductivity, show
that no such changes took place during the time of the impacts. The Lick Observatory
observations span a three week interval centered on the impacts and reproduce 8­months of
data obtained in 1991/1992. Upper limits on daily changes in the ionospheric conductivity
will be presented.
13

Global Imaging Campaign of Shoemaker­Levy 9 Impact Phenomena
M. W. Buie, L. H. Wasserman, R. L. Millis (Lowell Obs.),
W. B. Hubbard (U. Ariz.), H. L. Reitsema (Ball Aerospace and U. Ariz.),
F. Roques (Paris Obs.), A. Peyrot, M. Vignand (Obs. of Les Makes),
E. Dunham, C. Ford (NASA/Ames),
H. Hammel, J. Faust (MIT),
K. Meech, B. Patten (U. Hawaii),
J. Bell, III, (Cornell), D. Toublanc (Obs. Bordeaux),
R. Thompson (U. Wyoming)
We deployed imaging systems to ®ve sites for the time of the Comet Shoemaker­Levy 9
impacts on Jupiter. All sites used a combination of a high­speed optical CCD system with
similar broad and narrow­band ®lters.
Our coordinated effort covers from a few days prior to the ®rst impact to up to ®ve days
after the last impact. Most of the data consists of multi­spectral imaging that monitors the
morphology and development of the Jovian atmosphere as the impact week progressed.
Other high­speed (¸8 Hz) imaging sequences were taken from most sites at the time of
impact to search for transient phenomena related to the impact. Some plumes were recorded
through the 8900 #
A methane­band ®lter but most high­speed sequences showed no changes.
We will present a summary of the dataset as well as quanti®ed searches for satellite #ashes
and other limits on plume phenomena.
14

GALILEO INFRARED OBSERVATIONS OF THE SHOEMAKER­LEVY 9 G AND
R FIREBALLS AND SPLASH
R.W. Carlson, P.R. Weissman, J. Hui, M. Segura, W.D. Smythe,
K.H. Baines and T.V. Johnson
Jet Propulsion Laboratory
P. Drossart and T. Encrenaz
DESPA, Observatoire de Paris
F. Leader and R. Mehlman
Institute of Geophysics, UCLA
The Galileo spacecraft was fortuitously situated for a direct view of the impacts of comet
Shoemaker­Levy 9 in Jupiters atmosphere and measurements were recorded by the Near In­
frared Mapping Spectrometer (NIMS) instrument for several of the impact events. Seventeen
discrete wavelength channels were used between 0.7 to 5.0 microns, obtained with a time
resolution of 5 seconds. Two phases of the impact phenomona are found in the data: the
initial ®reball, which was evident for one minute, and subsequent fallback of impact ejecta
onto the atmosphere, starting six minutes after ®reball initiation.
Preliminary analysis of the G event data shows a ®reball appearing at 07:33:37 UT (as would
be observed from the Earth) with a temperature of 4000 K or greater and an effective source
diameter of 20 km or less. These spectra show absorption by methane and molecular hydrogen
whose strength place the ®reball in the upper troposphere, above the ammonia clouds. As
time progresses, the ®reball cools and the effective diameter increases about 2 km/sec. The
strength of the hydrogen and methane absorption decreases with time, indicating that the
radiating surface is rising supersonically. The ®reball appears to expand adiabatically, with
an adiabatic index of 1.2. After about 30 seconds, the spectra indicate a multiple temperature
or opacity structure.
A second phase of strong IR emission for the G event was detected beginning at 07:39:41
UT, which we interpret as impact ejecta, supersonically ejected in the ®reball and plume,
falling back upon the atmosphere. The resulting infrared emission steadily brightened over
2 minutes following its ®rst detection. The timing of the event implies a minimum ejection
velocity of 4 km/sec. Spectra of this splash phase suggest O­H and C­H stretch emission,
perhaps from shock heated water vapor and methane.
The R impact data show qualitatively similiar behavior, but with intensities reduced by a
factor of two to four.
15

Absolute Re#ectivity Spectra of Jupiter: 0.255 ­ 4.0¯m
Chanover, N.J., Beebe, R., and Simon, A.
New Mexico State University
Imaging data from the Hubble Space Telescope (HST) and several ground based observatories
are used to de®ne a reference absolute re#ectivity (I/F) spectrum of Jupiter for all latitudes.
The impact of Comet Shoemaker­Levy 9 (SL9) with Jupiter in July 1994 has resulted in a
tremendous amount of imaging data, primarily in the near­ultraviolet to near­infrared wave­
length regime. While these new data sets undoubtedly contain vast quantities of information
about Jupiter's atmosphere and its response to the comet fragment impacts, they were all
obtained from different telescopes, using different instruments, by different investigators.
Consequently, the calibration of all the data sets is not uniform. This makes comparisons
between data sets dif®cult, if not impossible. We have used high­resolution SL9 Impact
Campaign data from HST from 0.255 to 0.953 ¯m, along with near­infrared imaging data
taken at the Apache Point Observatory, Palomar Observatory, and the Infrared Telescope
Facility in July 1994, to create a reference spectrum of I/F as a function of wavelength for
all latitudes along the #undisturbed# central meridian of Jupiter. This spectrum will be made
available to all other SL9 investigators as a tool with which they may calibrate their own data.
An example of how this I/F spectrum can be used as a calibration tool will be illustrated with
February 1995 near­infrared imaging data taken at Apache Point.
16

Galileo Direct Imaging of Impacts K, N, and W
Clark R. Chapman and William J. Merline
Planetary Science Institute
Tucson, AZ USA
Kenneth Klaasen, Torrence Johnson, and Catherine Heffernan
Jet Propulsion Laboratory
Pasadena, CA USA
Michael J. S. Belton
NOAO
Tucson, AZ USA
Andrew Ingersoll
California Institute of Technology
Pasadena, CA USA
The Galileo Imaging Team
The Galileo camera obtained direct imaging data of the actual impact sites, showing the
luminous phenomena of bolide entry and early ®reball development for S­L 9 fragments
K and N (using a drift scanning mode) and for W (using time­lapse imaging). All three
bolides ± several­second­long initial #ashes ± were of comparable brightness (a few percent
of total Jupiter). But the K event showed a much brighter and longer­lasting ®reball (also,
possibly, a second bolide 10 seconds after the ®rst) than did the other two events, generally
consistent with the relative prominence of subsequent plume phenomena observed from
Earth. Later plume development for K and W was near or below our detection limits, as were
any secondary impacts during the intervals for which we returned data (with the exception
of the possible second, bright K event). We will present our latest lightcurves, calibrated
brightnesses, and compari­ sons with groundbased data for these events.
These Galileo data provide an excellent baseline for assessing Earthbased observations of
early phases of other S­L 9 impacts that were observed indirectly. In general, we are con®dent
that the onset of the luminous event we see is the bolide phase, and that precursor events
seen from Earth cannot be. Precursor events must generally be either the impact of coma or
much smaller, secondary projectiles (too faint to be seen by Galileo) in Jupiter's uppermost
atmosphere.
17

A Search for Small­Sized S­L 9 Impactors using a
Continuous­Readout CCD
Clark R. Chapman, Carol Neese, William J. Merline, Gail Schneller,
Stuart J. Weidenschilling, and Don R. Davis
Planetary Science Institute
Tucson, AZ USA
Philip Massey
NOAO
Tucson, AZ USA
We have performed sensitive observations of the Galilean satellites and Jupiter's morning
limb to search for #ashes of bolides by small, unnamed fragments of comet Shoemaker­Levy
9. To make the observations, we developed a technique of continuously reading the CCD
under a mask that covered all but Jupiter's equatorial plane. Continuous readout allows high
duty cycle and time resolution, while the mask prevents buildup of unwanted scattered­ and
sky­light. We cover Jupiter itself with a tiny neutral density ®lter. This avoids saturation
of Jupiter as we clock the chip at a rate slow enough to yield adequate S/N per line on the
satellites. Thus, we were able to observe Jupiter and the satellites simultaneously. The CCD
is rotated such that the clocking direction is tangent to Jupiter at the impact site (45 ffi S). In
this way, the charge from the streaked Jupiter does not overwrite the impact site, which is just
off the terminator. The resultant images are long trails of Jupiter and the satellites. Scans are
2 to 17 minutes in length, with time resolution of 25 ms per CCD line. Observations were
primarily done using a B ®lter, which we expected would give us the most sensitivity to the
hot bolide phase of the impacts.
The data were obtained on 8 nights at the Kitt Peak 2.1 m, including 3 nights before and
during the major impacts, and 5 nights afterwards (July 16, 17, 21, 27, 28, 29, 30, 31 UT). We
attempted to observe during impacts of fragments B and V. We encountered some problems
due to idiosyncrasies with the modi®ed instrumentation, but we were most affected by poor
weather. We obtained data during approximately 10% of our scheduled time. Preliminary
analysis shows no obvious impacts, although we may have observed the V #ash. Statistically
meaningful statements about smaller fragments must await more detailed analysis of the data.
18

Peculiarities of Spectral Evolution of SL9 Secondary Nuclei Impact Sites on Jupiter
Churyumov K.I., Tarashchuk V.P.
Astronomical Observatory of Kiev University, Ukraine
Prokof'eva V.V.
Crimean Astrophysical Observatory, Ukraine
Spectral observations of Jupiter were conducted at the Crimean Astrophysical Observatory
from July 2 to Aug. 30, 1994. Altogether ¸2000 spectra were obtained. The impacts
of large fragments in the Jupiter atmosphere caused powerful outbursts that led to complete
evaporation of a penetrated body as well as to heating of the surrounding atmosphere . Because
the comet fragment was surrounded by a coma that comprised products of evaporation of
the cometary matter and dust, then the region of penetration of cometary matter in the
atmosphere occupied much more space, than the place of the outburst and heating. Thus in
the impact sites there had to be luminosity of the evaporated cometary matter as the result of
the outburst yielding spectral lines that are absent in the Jupiter spectrum, and the spectrum
of the atmosphere itself had to undergo changes. The second effect of the penetration of
the fragment surrounded by the atmosphere could be appearance of certain emissions not
only in the site of the impact but at substantial distances from it as well. Analysis of the
spectra obtained in the Crimean Astrophysical Observatory allowed to detect a number of
peculiarities in the sites that appeared after the fall down of cometary fragments in the form
of dark spots. Spectral variability on temporal scales of tens of seconds was registered in the
sites of the impacts of the comet fragments 1 or 2 revolutions of Jupiter later after the impact
moments in the atomic lines and molecular bands of the Jupiter atmosphere (NH3 , CH4 ,
H2 and H). Intensities of the lines and bands also showed changes from spot to spot . Thus
in the spot spectrum that was caused by the fragment A impact observed 2.2 revolutions of
Jupiter , the biggest variability occurred in the region of the CH4 band. Variability in the
region of quadruple line of molecule H2 was registered when the impact site from fragment H
crossed the central meridian of Jupiter after 1.3 revolutions of the latter. Repeated variability
in the region of resonance line of Na at 589­589.6 nm as well as in the region of the band
of NH3 was registered several times. Occasionally variability in the lines of Li at 670.8 nm,
Ca at 643.9 nm and H2 at 656.3 nm occurred. Successive in time occurrence of variability
in different ranges of the spectrum are observed. And all this allows to suggest strati®cation
of outburst products and different excitation temperatures that gives the basis for seeking
possible regularities. Variability of Na with a big Doppler shift that was observed in several
cases is evidence of a complex picture of moving matter in the new spots. Preliminary
estimations of velocities of some jets of matter gave the value 300­600­1000 km/sec, here the
velocity vector had the direction both along the ray of the sight and in the opposite direction.
19

Hubble Space telescope Far­Ultraviolet Imaging of Jupiter During the Impacts of
Comet Shoemaker­Levy 9
John T. Clarke(1), Renee Prange(2), Gilda Ballester(1), John Trauger(3),
Robin Evans(3), Karl Stapelfeldt(3), and Wing Ip(4)
(1) University of Michigan
(2) CNRS ­ Universite de Paris, Orsay
(3) Jet Propulsion Laboratory
(4) Max Planck Institut fur Aeronomie
Hubble Space Telescope far­ultraviolet images of Jupiter during the Shoemaker­Levy 9
impacts show the impact regions darkening over 2­3 hours post­impact and darker and
more extended than at longer wavelengths, indicating UV­absorbing gases or aerosols more
extended, more absorbing, and at higher altitudes than the visible absorbers. Transient auroral
emissions were observed near the magnetic conjugate point of the K impact site just after
that impact. The global auroral activity was fainter than average during the impacts, while
a variable auroral emission feature was observed inside the southern auroral oval preceding
the Q1/Q2 impacts.
20

An Update on Imaging Observations from McDonald
Observatory
A. L. Cochran, B. J. Armosky, C. E. Pulliam, B. E. Clark,
W. D. Cochran, M. Frueh, D. F. Lester, L. Trafton (U. Texas),
Y. Kim (U. Maryland), C. Na (SWRI), W. Pryor (LASP)
During and shortly after the impact of SL9 into Jupiter, we obtained imaging data using the
0.8­m telescope and CCD detector, and 2.7­m telescope and HgCdTe detector at McDonald
Observatory. We present here further analysis of imaging observations which were obtained
during the impacts. We observed the impact of the R fragment into Jupiter and present optical
and IR light curves. In addition, we have continued to obtain CCD images of Jupiter since
the impact, including observations in January through March 1995. We will examine how
the impact sites have varied with time.
21

High Resolution CCD Spectroscopy of the Comet Shoemaker­Levy 9 Impact Locations
on Jupiter
William D. Cochran (University of Texas), Kevin H. Baines (JPL) Chan Na (SWRI), and
Wayne Pryor (LASP)
We used the McDonald Observatory 2.7m telescope 2Dcoud#e cross­dispersed echelle spectro­
graph to obtain high resolution (R = 60; 000) spectroscopy of Jupiter over the 4000­10000 #
A
region during and immediately after the impacts of the fragments of Comet Shoemaker­
Levy 9 on Jupiter. The spectrograph entrance aperture of 8:2 \Theta 1:2 arcsec allowed us to
obtain spatially resolved spectra of the impact spots and the surrounding undisturbed regions
of the Jovian atmosphere. We model the observed H 2 , NH 3 and CH 4 lines in order to derive
the altitude and optical thickness of the stratospheric haze caused by the impacts as a function
of wavelength and of time. We also present results of a search for absorption by H 2 O in the
visible spectrum.
22

Jovian Acoustics and Gravity Waves
Michael D. Collins and B. Edward McDonald
Naval Research Laboratory, Washington, DC 20375
W. A. Kuperman
Scripps Institution of Oceanography, La Jolla, CA 92093
William L. Siegmann
Rensselaer Polytechnic Institute, Troy, NY 12180
We have modeled the propagation of acoustic and gravity waves from the impact sites of
Comet Shoemaker­Levy 9 using a three­dimensional technique that accounts for spatial
variations in sound speed, BV frequency, density, and wind velocity. The zonal winds act
as horizontal waveguides that pinch the propagating energy into beams that remain coherent
all the way around Jupiter. Since the energy in the beams is enhanced by about an order
of magnitude, they indicate the most favorable locations to look for evidence of waves far
from the impact sites. It has been established that gravity waves appear near the impact sites
in HST images [H. B. Hammel et al., Science 267, 1288­1296 (1995)]. Some of the HST
images also show evidence of an outward moving front that is at a location consistent with
the group speed of an acoustic wave.
23

Palomar Mid­infrared Spectroscopic Observations of Comet Shoemaker­Levy 9
Impact Sites
B.J. Conrath
NASA/GSFC
P.J. Gierasch, T. Hayward, C. McGhee,
P.D., Nicholson, & J. Van Cleve
Cornell University
Post impact spectra between 8 and 14 micrometers of the Shoemaker­Levy 9 impact sites
and surrounding regions, acquired with SpectroCam­10 on the Hale telescope, are analyzed.
Both direct modeling and inversion techniques are used to infer information on Jovian atmo­
spheric composition and thermal structure from these observations. In the days immediately
following the impacts, over areas with diameters of approximately 2000 km centered on the
sites, stratospheric temperatures are found to be elevated relative to adjacent, unperturbed
areas. A parameterized retrieval of the vertical distributions of ammonia and phosphene indi­
cate that these gases are substantially enhanced in the stratosphere, and strong stratospheric
emission from ethylene is also observed. The temporal behavior of the retrieved atmospheric
parameters associated with sites K and L over a 5­day post impact period is summarized.
Stratospheric aerosol opacity in the 10­micrometer region is presented. Finally, implications
of these results concerning the radiative and dynamical processes associated with the sites
are considered.
24

Near­Infrared Imaging Spectroscopy of the Impacts of SL9 Fragments C, D, G, K, N,
R, V, and W with Jupiter
D. Crisp (JPL) and V. Meadows (NRC/JPL)
We used the InfraRed Imaging Spectrometer (IRIS) on the Anglo­Australian Telescope to
observe the collisions of 8 of the Comet Shoemaker­Levy 9 fragments with Jupiter. We also
monitored the evolution of the impact sites from 16 to 23 July, 1994. Spatially­resolved
K±band (1.98±2.38 ¯m) spectral image cubes were collected during the C, D, G, K, R and W
impact events. An H±band (1.4±1.8 ¯m) image cube was also collected during the Fragment
K impact event. Fast­rate 2.35 ¯m ®lter photometry was used to monitor the impacts of
fragments N and V. The IRIS image cubes provide a spectral resolution of 300 (–=\Delta–), a
spatial resolution of 0.6 00 =pixel, and a temporal resolution of ¸ 2 minutes or less (¸ 30s for
fragment K). These observations therefore provide detailed descriptions of the impact events,
from the time the fragments entered the Jovian atmosphere, until their collapsed plumes
crossed the limb and traversed the day side of Jupiter. Many of the AAT/IRIS observations
were simultaneous with Galileo or HST observations, facilitating comparisons among these
data sets. Each event was ®rst detected as one or more faint #ashes (precursors). The G and
K precursors were ®rst detected 0.5 to 1.5 minutes before the impact ®reball was observed
by instruments on the Galileo spacecraft. We acquired spectra of both the ®rst (entry)
precursors, and second (®reball) precursors for fragments C and K. The ®rst precursors'
spectra were dominated by blue continuum emission. The second precursors' spectra were
much redder, and exhibited enhanced CH 4 absorption features at 2.20, 2.32 and 2.70 ¯m as
they faded. These precursors were followed 5±10 minutes later by a much brighter #main
#ash" which peaked 8­15 minutes after the second precursor. The main #ashes were also
initially dominated by continuum emission with color temperatures decreasing from ?500
to ¸ 230 K. Weak CO absorption at wavelengths ? 2:3 ¯m was also occasionally detected
during the initial phase of the main #ashes. As the intensities of the main #ashes peaked
and started to fade, our spectra revealed strong line emission by CO (– ? 2:29¯m), NH 3
(2.03 ¯m), CH 4 (2.20 ¯m) and H 2 O (2.0 and 2.3 ¯m). H 2 (2.122 ¯m) quadrupole emission
was also detected during the decay of the the C and D #ashes. The CO line emission indicated
rotational temperatures ? 2000 K. Thirty minutes after each impact, there was little or no
evidence of CO, H 2 , or H 2 O absorption and emission in our spectra of the impact clouds.
The initial dominance of continuum emission over line emission in the main infrared #ashes
suggests emission from hot particulates rather than gases in the rising impact plume. CO
seen in absorption in the initial phases of the main #ash may have been transported upwards
in the expanding, cooling, ®reball. The hot CO, NH 3 and H 2 O seen during the decay of the
main #ash are indicative of a ‘splashback' event as the plume collapsed. Emission rates,
and an estimate of the total abundance of emitting molecules will be presented. A radiative
transfer model was also used to estimate the altitude and particle properties of the impact
clouds. The spectra of the impact sites several hours after impact can be simulated by an
optically­thin (Ü ? 0.25) cloud consisting of small (¸ 0.25 ¯m),re#ective (! ? 0:97) particles,
located above the 1 mbar level.
25

THE R­IMPACT: FLASHES AND FIREBALLS, WHAT DID WE SEE?
Imke de Pater, P. Nicholson, K. Zahnle, J. R. Graham, G. Jernigan,
T. Hayward, G. Neugebauer, K. Matthews, A. Weinberger,
M. Liu, M. Brown
The R­impact was observed at 2.3, 3.2 and 4.5 ¯m with the Keck and Palomar telescopes
respectively. We will show a comparison of the various datasets, and compare the event
times with the timing of the #ashes observed by the Galileo spacecraft. Based upon the
viewing geometry and timing of the various events, we suggest the following quantitative
scenario for the sequence of events. Given the timing of Flash 1, it must be related to the
comet's entry into Jupiter's atmosphere. The time evolution of the #ash is very similar at
the three wavelengths; it lasts for about 40 seconds and the #ux ratios are consistent with a
temperature of ¸ 1000 K. We suggest it is the meteor trail seen about 400 km above Jupiter's
cloud deck. The duration of the trail is indicative of the cooling time of the heated trail high
up in the atmosphere, or it may suggest the comet fragment to be broken up into a string of
material about 2000 km in extent. Flash 2, starting approximately one minute after the ®rst
#ash appeared, has a rapid rise time at all three wavelengths. This #ash is roughly 20 times
brighter at 4.5 ¯m than at 2.3 ¯m. We attribute the second #ash to thermal emission from
the ®reball rising above the limb of Jupiter. The decay, presumably indicative of adiabatic
cooling, is slow at 2.3 ¯m (¸ 3 min), but quite fast at 4.5 ¯m (¸ 30 sec). This is just
opposite to what one would expect for a slowly cooling ®reball, and is not yet explained.
Galileo saw its ®rst #ash ¸ 15 seconds before Keck and Palomar saw the second #ash. Since
Galileo NIMS had a direct view of the impact site, it was able to see the ®reball well before
Earth­based observers.
The evolution of the third and brightest #are is similar at all three wavelengths. We attribute
this #are to a high atmospheric temperature (T ú 600 K) induced by the release of gravita­
tional potential energy by ®reball material raining down onto the atmosphere. The beginning
of the #are was also seen by Galileo. A fourth peak seen 10 min after the main #are is similar
to those reported from the H and L impacts. At optical wavelengths, HST observed several
plumes (A, E, G and W) in re#ected sunlight. Although the infrared #ares and HST plumes
are generally seen at the same time, we believe that they represent different aspects of the
same elephant.
26

THE OUTBURST OF JUPITER'S SYNCHROTRON RADIATION FOLLOWING
THE IMPACT OF COMET P/SHOEMAKER­LEVY 9
Imke de Pater, C. Heiles, M. Wong, F. van der Tak, R. Millan (UCBerkeley),
R.J. Maddalena (NRAO Green Bank) M.K. Bird, O. Funke (Univ. of Bonn),
J. Neidhoefer (MPI Bonn), R. M. Price, M. Kesteven (CSIRO, Australia),
M.J. Klein, S.J. Bolton (JPL), R.S. Foster (NRL),
S. Sukumar (DRAO,Penticton), R.G. Strom (NFRA, Dwingeloo),
R.S. LePoole (Leiden Univ., Netherlands),
R.W. Hunstead(Sydney Univ, Australia)
Jupiter's microwave emission was observed since well before the SL9 impact by many
different telescopes. We present the total intensity data at wavelengths between 6 and 90 cm
(an improved subset of the data published by de Pater et al., 1995, submitted to Science, Dec.
1994). The #ux density increased, depending on wavelength, by 10±45% during the week
of impacts, and the radio spectrum hardened. Following the week of cometary impacts, the
#ux density began to subside at all wavelengths, while the spectrum continued to harden.
The #ux density was still declining in February 1995. We suggest that the increase in #ux
density is caused by energization of the resident particle population. The evolution of the
radio spectrum after the SL9 encounter may be caused by an enhancement in the amount
of magnetospheric dust, due to the break­up and #outgassing# of the comet. The radio #ux
density measurements may be the only dataset sensitive enough to determine the amount of
cometary dust injected into the magnetosphere. As such, the radio data provide a unique
probe of radial diffusion in Jupiter's magnetosphere.
We show the time evolution of the brightness distribution of the radio emisison as observed
by the VLA and the WSRT telescopes. A preliminary reduction and analysis of the VLA
images shows, crudely, the following phenomena: a) The enhancement in Jupiter's #ux
density is, initially, dominated by an enhancement in the magnetic equator at longitudes
– III ú 160 \Gamma 260 ffi , in the socalled active sector. b) There is no obvious connection
with the impact sites or times. c) The radiation peak which brightened is often (but not
necessarily always) displaced inwards, closer to Jupiter. This suggests that radial diffusion
plays some role and that (part of) the increase in the radio emission can be explained through
electron energization by conservation of the ®rst adiabatic invariant. d) The impact­induced
asymmetry between the peaks lasts for less than 4 days, as expected from the energy dependent
drift period of the energetic particles.
27

Flaring­up Splitting of SL9 after Encounter with Jupiter
E.M. Drobyshevski
Ioffe Physico­Technical Inst.
194021 St.Petersburg, Russia
E­mail: emdrob@drob.pti.spb.su
Assumption of tidal disruption of SL­9 contradicts numerous data: it lasted months after the
perijove passage, the fragments' velocity depended on their size, etc. One can hardly explain
why this breakup should be distinguished from rather common breakups of other comets far
from planets. Such breakups are believed to be caused by the inner energetics of comets
and they correlate with a #are comet activity which, in its turn, is tied somehow with the
Solar activity. The inner energy sources of SL­9 were triggered likely by the same cause that
initiates the sources in comets when they cross sector structure boundaries in the Solar wind,
viz. by electric current #own through the SL­9 nucleus when it found its way deep into the
Jovian magnetosphere and 100 kV e.m.f. was induced across the nucleus (recall that 400
kV e.m.f. induced across Io generates 5 MA current). What an energy source is required
for: (1) being initiated by the current, (2) being capable of imparting 1 m/s velocity to 1
km fragments, and (3) being active in them during months? The only option reported thus
far could be the dirty ice saturated with its electrolysis products and so capable of burning
and even detonation when some additional energy is supplied. This hypothesis is proved
to be fruitful as for numerous minor bodies' origin understanding, explanation of SP comet
energetics and chemistry, including their distant outbursts etc, and for explanation why in
SL­9 fragments and in plumes originated after their infall onto Jupiter no water traces were
detected. Here water could evaporate due to electrolized ice outburning before discovery of
SL­9 and its hitting Jupiter.
28

Lithium Presence in Impact Plume Favors the Planetary Origin of SL9
E.M. Drobyshevski
Ioffe Physico­Technical Inst.
194021 St.Petersburg, Russia
E­mail: emdrob@drob.pti.spb.su
Lithium is one of the less abundant elements. Its detection in SL­9 L fragment plume has
risen a number of questions. Two ones are main: (1) why was Li observed never before, even
it Sun­grazers, and (2) why was it localized in some portion of the SL­9 ice only? The great
localized Li overabundance can be hardly explained in framework of standard hypotheses on
the primary cometary grains' formation in turbulized gaseous medium. The New Eruptive
Cosmogony (NEC) of minor bodies considers the SP comet nuclei to be fragments of icy
envelopes of Ganymede­type bodies. Such ice accumulates its electrolysis products 2H2+O2
as the solid­state solution. On reaching their 15­20 wt.can explode being initiated by the
meteoroid impact. The local Li concentration in comet ice can be understood from the NEC
standpoint if one takes into account geochemical processes occurring inevitably in parent
planetary bodies. These processes involve at ®rst an igneous concentration of Li in acid
and sedimentary rocks of the rocky core crust and, afterwards, the aqueous washing out
and electrolytic transport of Li compounds into the water­ice mantle. These compounds are
accumulated there in the last salt brine resting after the water freezing (or evaporation). By
this way Li can be accumulated in the form of separate ore nests near the rocky core surface
and in the ice envelope. Thus, in the context of NEC, it is not surprising that Li occurs to be
concentrated only in a small part of far from all the cometary nuclei.
29

Changes in Flux and Brightness of Jupiter's Synchrotron Radiation
During and After SL­9 Impacts
George A. Dulk 1;2 , Yolande Leblanc 1 and Richard W. Hunstead 3
1 DESPA, Observatoire de Paris, 92195 Meudon, France.
2 University of Colorado, Boulder, CO 80309­0391, U.S.A.
3 University of Sydney, Sydney, NSW 2006, Australia
The Australia Telescope and the Molonglo Observatory Synthesis Telescope were used to
observe Jupiter before, during and after the impacts of SL­9. Integrated #ux and 2­D images
of Jupiter's synchrotron radiation belts were produced with a resolution of 4 00 \Theta 18 00 at 13 cm,
7 00 \Theta 30 00 at 22 cm and 43 00 \Theta 200 00 at 36 cm.
Starting about one day after the ®rst impact and reaching a peak soon after the time of the last
impact, the integrated #ux density increased by approximately 30% at 13 cm, 25% at 22 cm
and 40% at 36 cm.
Before the impacts, at 22 cm, the images were similar to those observed at 20 cm by the
VLA 12 years earlier at the same longitudes. During and after the SL­9 impacts the belts
were brighter than before, and remained bright at least 12 days after the ®rst impact. The
increase in brightness was almost entirely con®ned to one hemisphere, between about 100deg
and 240deg longitude. The changes in the belts at 13 cm were similar to those observed at
22 cm. The new asymmetry implies the existence of a new or newly­accelerated population
of relativistic electrons con®ned to a portion of the magnetic ®eld of the planet.
30

Abundances of NH 3 and CS 2 and Upper Limits of H 2 S, PH 3 , C 2 H 2 , and C 2 H 6 Above
the SL9/G Impact Site
S.G. Edgington and S.K. Atreya
Department of Atmospheric, Oceanic and Space Sciences
University of Michigan, Ann Arbor, MI 48109­2143
L.M. Trafton Department of Astronomy
University of Texas, Austin, TX, 78712
J.J. Caldwell
Institute for Space and Terrestrial Science
Space Astrophysics Lab., 4850 Keele St., North York, ON, M3J 3K1
K.S. Noll and H.A. Weaver
Space Telescope Science Institute
3700 San Martin Drive, Baltimore, MD 21218
Two­stream radiative transfer models allowing for anisotropic scattering were used to ®t the
July 18, 1994 Hubble Space Telescope/Faint Object Spectrograph 170±230 nm observations
of Jupiter above the G­impact site. Aerosol unit optical depths were varied to occur within
0.3±10 mb and the refractive index was varied to account for a range of possible aerosol
materials. The models yield abundances for NH 3 and CS 2 of (0:25 \Gamma 1:3) \Theta 10 16 cm \Gamma2
and (0:43 \Gamma 1:1) \Theta 10 15 cm \Gamma2 , respectively, at the unit optical depth level in the part of
the atmosphere above the G­impact site. The upper limits of H 2 S, PH 3 , C 2 H 2 , and C 6 H 6
at the unit optical depth level are 1:2 \Theta 10 16 cm \Gamma2 , 3:3 \Theta 10 14 cm \Gamma2 , 1:0 \Theta 10 17 cm \Gamma2 , and
3:3 \Theta 10 14 cm \Gamma2 , respectively.
31

Near­Infrared Spectroscopy of Jupiter at the Time of Comet Shoemaker­Levy 9
Impacts: Emissions of CH 4 , H +
3 AND H 2
Th.Encrenaz and P.Drossart (DESPA, Obs.Paris), R.Schulz and J.A.Stuewe (MPIA Lindau),
G.Wiedemann (ESO, Garching), and J.Crovisier (ARPEGES, Obs.Paris)
Near­infrared emissions of the SL9 impact sites of Jupiter have been recorded on July 16­18,
1994, using the IRSPEC imaging spectrometer at the 3.5m NTT of ESO (La Silla, Chile). A
very strong emission of methane was recorded between 3.50 and 3.56 microns, shortly after
impact H, showing evidence for a strong and temporary increase of the Jovian stratospheric
temperature. Emissions of H2 (2.12 microns) and H3+ (3.53 microns) werew detected
above some of the impact sites, several hours after the impacts. A strong continuum was
also detected at 2.12 microns over the impact sites, presumably due to intense scattering of
re#ected sunlight by high­altitude stratospheric haze.
32

Infrared Heterodyne Observations of NH 3 and C 2 H 6 After The Collision of Comet
P/Shoemaker­Levy 9 With Jupiter
K. E. Fast(UMD), T. Kostiuk(NASA/GSFC), F. Espenak(NASA/GSFC), D. Zipoy(UMD),
D. Buhl(NASA/GSFC), T. A. Livengood(UMD), G. Bjoraker(NASA/GSFC),
P. Romani(NASA/GSFC), J. Goldstein(SI/NASM)
Spectra were taken with the Goddard Infrared Heterodyne Spectrometer during the NASA/IRTF
observing campaign of the collision of comet P/Shoemaker­Levy9 with Jupiter. Stratospheric
NH3 emission was observed at a number of impact regions days after the collisions. The true
line shapes, obtainable with this instrument, indicate NH3 emission from between 1 and 10
mbar at the impact regions. Stratospheric C2H6 emission was also observed along the impact
latitude. Preliminary analysis and conclusions will be presented, along with the results of
follow­up measurements made in March '95.
33

HST Observations of Mg + and Dust in Outburst
from Comet P/Shoemaker­Levy 9
P. D. Feldman (JHU), H. A. Weaver (STScI), D. C. Boice (SWRI),
S. A. Stern (SWRI), M. A. McGrath (STScI)
Ultraviolet spectroscopy of SL9 was performed on three different occasions between July
1993 and July 1994 (Weaver et al. 1994, 1995) using the Faint Object Spectrograph (FOS)
of the Hubble Space Telescope with the primary objective to search for the hydroxyl radical
(OH) as a tracer of water ice in the SL9 fragments. Although OH was not detected during
these spectroscopic investigations, strong Mg II emission near 2800 #
A was observed in the
vicinity of fragment G during a very short outburst on 14 July 1994. The emission decayed
with a time constant of about one minute. Approximately 18 minutes after the Mg + outburst,
the continuum spectrum of dust scattered sunlight also changed dramatically, increasing by
more than a factor of three over an 8 minute period and then relaxing back to the quiescent
level less than 20 minutes after the start of the outburst. The enhanced continuum also
exhibited a strong reddening, ¸13% per 100 #
A, compared with a #at solar spectrum for the
quiescent dust.
The two lines in the Mg II doublet are partially resolved in the HST spectrum, indicating
that the emission was strongly peaked within a spatial region of ¸1000 km and suggest­
ing ion sputtering of the dust surrounding the fragment as a possible source mechanism.
This structure also excludes the possibility of the Mg II emission arising from a terrestrial
atmospheric source. At the time of these observations the comet was 3:8 \Theta 10 6 km away
from Jupiter, which is consistent with the expected location of the magnetopause. Neutral
Mg was not detected, although the strong Mg I resonance transition at 2853 #
A was in our
bandpass. Quantitative estimates of the amount of mass released in these two events and
physical constraints on possible source and loss mechanisms will be presented.
References:
H. A. Weaver et al., 1994. Science 263, 787.
H. A. Weaver et al., 1995. Science 267, 1282.
34

Calibrating groundbased observations of the SL­9 impact clouds with small telescopes
using the HST WFPC2 images as 'ground truth'
Daniel Fischer
Planetary and Cometary Observers Group
Germany
Carlos Hernandez
Association of Lunar and Planetary Observers
U.S.A.
While the images from the HST have undoubtedly the highest angular resolution, the temporal
coverage of the evolution of the impact clouds on Jupiter is very uneven and displays large
gaps; furthermore, due to sun proximity concerns no images could be obtained for many
months after late August. In contrast, a tremendous network of dedicated amateur observers
was following the developments on Jupiter almost minute by minute for months, and it has
been suggested that these CCD images and drawings with small telescopes form a valuable
resource for our understanding of both the impact physics and the wind motions on Jupiter
(Limaye, AGU Fall Meeting 1994). We are comparing archival HST WFPC2 images with
simultaneous amateur observations to determine the precision and general reliability of the
latter and to arrive at rules one should follow in interpreting them.
35

Near­IR Spectral Images from WIRO using the U. of Rochester Array of Jupiter
during Comet Shoemaker­Levy 9 Impact Interval
Brendan Fisher, Judith Pipher, William Forrest, & Eric Howard
U. Rochester
Chick Woodward & Bob Howell
U. of Wyoming
Bob Gehrz
U. of Minnesota
The University of Rochester Infrared group, in collaboration with the University of Wyoming
and the University of Minnesota collected a large set of near­IR image data of Jupiter in July
1994 from the 2.3­m telescope at the Wyoming Infrared Telescope Facility using a 256x256
InSb array. We obtained a large number of images in the 1.6­2.3 and 2.9­4.0 micron bands
with two CVFs with a spectral resolution of 2was imaged in the J, H, K, L', and M' (4.67)
bands. The B, U, and V impacts were observed but no effects were seen in the preliminary
analysis. Other sites were observed after the impacts as permitted. We have observed an
unusual spectral emission in the Q1 spot near 3.35 microns during its ®rst transit that was
not visible in other spots visible on the disk at the same time. Other preliminary results and
progress will be presented along with a summary of our spectral­temporal coverage.
36

Optical spectroscopy of atomic emission from the L and Q 1 impacts on Jupiter
A. Fitzsimmons, J. E. Little
Queen's University of Belfast
Belfast, UK
P.J.Andrews, R.Catchpole, N.Walton
Royal Greenwich Observatory
Cambridge, UK
I.P.Williams
Queen Mary & West®eld College
London, UK
During the impacts of fragments Lwe observed the associated low­resolution optical spectrum
at wavelengths 350nm­670nm. The spectra demonstrate the existence of neutral Na, Fe, Mg,
Ca and Mn in the plume, which we identify as remnants of the vapourised bolide. The
following night we also obtained high­resolution spectra of Na D doublet emission resulting
from the impact of fragment Q1. We have constructed models that are able to explain the
observed emission line intensities. We will present our latest modelling of the temperatures
and column densities of the various atomic species observed.
37

Model Calculations of Aerosol Particle Mean Radius and Number Density Formed in
Comet SL­9 Fireballs
A.J. Friedson, R.A. West, and K.H. Baines
JPL/Caltech
We present calculations of the mean radius and number density of particles condensing in a
®reball expanding along a moist adiabat, for various assumed compositions for the aerosol.
The initial mass of condensible vapor in the ®reball is an important parameter. Relatively
volatile species such as ammoniatend to form particles less than a few tenths of a micrometer
in radius; refractory species generally produce particle radii greater than 1 micrometer.
We estimate the optical depth of plumes seen above Jupiter's limb at visible and near­IR
wavelengths for different cases of composition and vapor mixing ratio.
38

Long Term Monitoring of Jupiter's Synchrotron
Radiation with the Nan#cay Radiotelescope at the Time
of the Collision with Comet P/Shoemaker­Levy 9
Patrick H. M. GALOPEAU, Eric GERARD
and Alain LECACHEUX
Observatoire de Paris, URA CNRS 1757
5, Place Jules Janssen, F­92195 Meudon Principal Cedex, France
The jovian synchrotron emission has been monitored with the Nan#cay radiotelescope at 1410
MHz, 1666 MHz and 3300 MHz from 1 April 1994 to 23 February 1995. Between 13 July
and 26 July 1994, we have observed a 20 % increase of the non­thermal #ux at 1410 and
1666 MHz but a 40 % increase at 3300 MHz. In February 1995, the non­thermal #uxes at
the three frequencies are still 50 % above the pre­encounter levels. There is no signi®cant
change of the beaming curve during the pre­ and post collision periods. The possibility of
changes in the degree of linear polarization and source size are investigated.
39

Multi­Filter Optical Observations of SL9 Impact Spots on Jupiter
Galen Gisler
Los Alamos National Laboratory
Bryan Laubscher
Amparo Corporation
Nancy Chabot
Los Alamos National Laboratory
Kirsten Boudreaux and Randy Grashuis
University of New Mexico
The 24­inch telescope of the Capilla Peak Observatory of the University of New Mexico
was used to observe Jupiter during the week of the impact of comet Shoemaker­Levy­9
onto Jupiter. Images of Jupiter were obtained in several narrow­band interference ®lters. The
opacity in the spots is calculated as a function of wavelength, and used to infer the distribution
of the debris with altitude. The debris is found to reside very high in the aerosol layer above
the visible cloudtops.
40

IRTF Impact Imaging Results
J. D. Goguen, A. J. Friedson, G. S. Orton
JPL
The IRTF Science Team
Highlights of the IRTF results are published in the special issues of Science and Geophysical
Research Letters. This poster will include some of those highlights pertaining to impact
imaging of precursors and plume evolution, but will focus on aspects of the impact data set
not previously reported. The IRTF acquired useful data for the B, C, F, G, and R impacts.
41

Mid­IR Spectroscopy and NH 3 and HCN Images of K Impact Site
C.A. Grif®th (N. Arizona U.), B. B# ezard (Paris­Meudon Obs.), D. Kelly (U. Texas),
J. Lacy (U. Texas), T. Greathouse (N. Arizona U.), & G. Orton (NASA/JPL)
We present observations of SL9 comet's crash that were taken at the NASA Infrared Tele­
scope Facility (Mauna Kea, Hawaii), with the U. Texas mid­infrared echelle spectrometer
IRSHELL. This is a 21 spatial by 64 spectral array, with a step size of 0.8 arcsec. Observa­
tions were conducted at a spectral resolution of ¸15000. We detected ammonia emission at
wavelengths of 908 and 948 cm \Gamma1 over the K impact site at several times after impact: 20
hours, 5 days and 10 days later. HCN emission was observed 10 days after impact. With each
detection, both the NH 3 and HCN emission lines were imaged with a resolution of 1.4 arcsec
over a 17 by 6 arcsec swath centered on the K impact site. At the impact site the ammonia
lines were best interpreted with a distribution mostly above 10 mbar, having a column density
of 2 \Gamma 5 \Theta 10 17 molecules cm \Gamma2 above 40 mbar. We have derived both vertical information
and horizontal distributions for NH 3 and HCN. The inferred HCN abundance is sensitive to
the temperature pro®le, which was derived from C 2 H 2 and CH 4 line spectra (see poster by
B#ezard et al.). We observe the ammonia spreading out over a larger horizontal area with time.
Its distribution is depleted with time as a function of altitude, consistent with photochemical
destruction (Moses et al.1995). The vertical distribution of the NH 3 appears consistent with
dynamical models of the explosion (K. Zahnle, priv. comm.).
42

Microwave Imaging of Jupiter's Troposphere
During Impact with Comet P/Shoemaker­Levy 9
A. W. Grossman and S. M. White (Univ. of Maryland),
D. O. Muhleman and M. A. Gurwell (Caltech)
We report on the results from a world­wide campaign to acquire high­resolution, microwave
images of thermal emission from Jupiter's troposphere, before, during, and after impact with
fragments of comet P/Shoemaker­Levy 9. Interferometric observations were obtained at
wavelengths of 3cm and 6cm from the Very Large Array (VLA) and the Australia Telescope
(AT). At these wavelengths, gaseous ammonia is the primary source of opacity in Jupiter's
troposphere. It is also the principal condensate. Therefore changes in brightness temperature
are indicative of changes in ammonia abundance and relative humidity. The corresponding
weighting functions at these wavelengths probe the pressure levels 1­10 bars, below the
optically thick cloud deck.
Preliminary images at centimeter wavelength clearly show the expected zone­belt structure.
Two prominent features in the radio data are a distinctly bright band at the position of the
North Equatorial Belt (NEB) and a corresponding dark band at the position of the Equatorial
Zone (EZ). The 30K brightness temperature difference (at a wavelength of 6cm) between
these two regions, interpreted within the context of a radiative transfer model, requires that
the NH 3 mixing ratio decrease by a factor of two from the EZ to the NEB.
At the latitude of the impacts there are no discernible features at a level of 3­5K within
a 3000­5000 km beam­width. This constrains the change in the average NH 3 abundance
within this region to value of less than 10%, although larger changes in NH 3 abundance are
allowed in a smaller region. These limits may constrain the depth of penetration of the comet
fragments. Work continues to reduce the full complement of data taken during the week of
impacts and to improve the sensitivity and quality of the images.
43

EUVE Observations During the Comet Shoemaker Levy/9 Impacts
D. T. Hall
Center for Astrophysical Sciences
The Johns Hopkins University
Baltimore, MD 21218
G. R. Gladstone
Southwest Research Institute
6220 Culebra Road
San Antonio, TX 78228
The Extreme Ultraviolet Explorer satellite conducted extensive observations of the Jupiter
system before, during and after the arrival of the fragments of Comet Shoemaker­Levy/9.
Approximately 500,000 seconds of data were acquired in 1994. Previous observations of
the Jupiter system revealed that electron­impact generated line emission from oxygen and
sulfur ions resident in the Io plasma torus dominate the spectrum in the 37 to 73 nm range;
these emissions dimmed 30±50% during the impact events, and no new torus emissions
were detected. Because no similar reduction in Io torus output was observed at longer
wavelengths, the EUV luminosity reduction was very likely caused by a reduction in the
population of high­energy Io torus electrons (Ee ? 17 eV), which excite a large fraction
of the EUV emissions but a comparatively small fraction of longer wavelength emissions.
Following several of the impacts, EUVE detected transient brightening events in planetary
atomic helium resonance emission, HeI 58.4 nm, which was barely detected before the
impacts (I ú 3 Rayleigh), although it was measured by the Voyager spacecraft during 1979
(I = 4 Rayleigh). After the impact of fragment Q1 the HeI 58.4 nm brightness increased to
25±70 Rayleigh, or about 8±25 times the average brightness of Jupiter before the impacts.
The brightening events indicate that the impacts lifted substantial amounts atomic helium
to high altitudes in Jupiter's atmosphere, above much of the molecular hydrogen gas that
normally absorbs HeI 58.4 nm photons emitted from lower altitudes.
44

An Interpretation of Calar Alto Ground­Based Infrared Lightcurves
D.P. Hamilton (Max Planck Institut f #
ur Kernphysik, Heidelberg),
T.M. Herbst (Max Planck Institut f #
ur Astronomie, Heidelberg),
H. B# ohnhardt (Universitats­Sternwarte, M# unchen),
J.L. Ortiz (Instituto de Astro®sica de Andalucia, Granada),
A. Richichi (Max Planck Institut f #
ur Astronomie, Heidelberg), and
G. Calamai (Osservatorio Astro®sico di Arcetri, Italy).
We describe 2.3¯m and 3.1¯m data obtained at Calar Alto's 3.5­meter and 2.2­meter tele­
scopes during the impacts of Shoemaker­Levy 9 fragments A,E,H,L,Q1,Q2, and S with
Jupiter. All of these impact events produced signi®cant brightening of the jovian disk lasting
on the order of 10­20 minutes. These #main events# are the brightest features seen by infrared
and optical telescopes. Five of the impacts ­ those for fragments A,H,L,Q1, and Q2 ­ show
fainter precursor features which begin roughly ®ve minutes before the main events. For
impacts H and L, there are additional features at still earlier times. These sharp H and L ®rst
precursors rise, peak, and begin to fade before the onset of the signals detected by Galileo's
PPR instrument.
Signals from Jupiter might be expected from several stages of each impact including the
bolide entry into the jovian atmosphere, the rising impact shock wave and ®reball, re#ected
sunlight and thermal emission from the ejecta plume, reimpacting ejecta, and the rotation of
the glowing impact site into view. Interpretations are dif®cult due to the far­side location
of the impact sites and the changing geometry from one impact to the next. We present
our infrared lightcurves and attempt to interpret speci®c features in these curves in terms of
physical mechanisms.
45

Observations of the Dynamic Response of Jupiter's Atmosphere to the Impact of
Comet Shoemaker­Levy 9
J. Harrington, T. E. Dowling, C. M. Santori, H. B. Hammel, J. R. Mills (MIT),
A. P. Ingersoll (Caltech), R. F. Beebe (NMSU),
G. S. Orton, P. Chodas, D. Yeomans (JPL)
We present HST images of impact sites A, E, G, Q1, and R during their ®rst planetary
rotations (Hammel et al., Science, 1995). They each show one or two rings expanding from
the impact site and located above the atmospheric methane (P ! 380 mbar). Outer ring
radius varies linearly with time at a velocity of 454\Sigma20 m/s. The inner ring ®t gives 353\Sigma83
m/s. We interpret the rings as signatures of stratospheric inertia­gravity waves, with the outer
ring speed indicating a deformation radius of ¸1,800 km, consistent with current estimates.
The observed outer ring speed corresponds to the ¸400 m/s waves in the 16±69 mbar layer
of our pre­impact model (Harrington et al., Nature, 1994). Tropospheric IG wave speeds
in this model and that of (Ingersoll et al., GRL, 1994) are closer to 100 m/s. The latter
model predicted that a water­cloud wavefront (5 bars pressure) would affect the stratosphere.
However, the observed outer wave moves too fast to be driven from the troposphere; this is
also likely true of the inner wave. One possibility for faster tropospheric waves in that model
is a water abundance many times that of estimates based on solar composition.
This work was supported by NASA.
46

Jeans Instability in Comet Shoemaker±Levy 9
Joseph M. Hahn and Terrence W. Rettig
University of Notre Dame
Recent N±body simulations of a tidally disrupted comet (e.g., Asphaug and Benz 1994)
indicate that gravitational instabilities may be responsible for the appearance of comet
Shoemaker±Levy 9. These models show that if the comet catastrophically fragmented into
N AE 20 parts, the particles can gravitationally recondense into n fragments a few hours
after periapse. Since n depends sensitively on the parent comet's density ae, catastrophic
disruption of a comet provides a sensitive measure of its density. We show, via a tidal
impulse model, that if tides disrupted S±L 9 into a cloud of many particles, the resulting
cloud is Jeans unstable and condenses to ¸ 20 fragments about 10 hours after periapse for a
comet density ae ¸ 0:6 gm/cm 3 . Whether these instabilities condense to compact objects or
remain as extended swarms depends upon the rate at which particle collisions might damp
their random motions. We also show that a varicose instability in an incompressible medium
cannot account for S±L 9's twenty fragments, for this instability quickly breaks up a tidally
elongated cylinder into only n ¸ O(1 or 2) fragments. These results indicate that a particle
gas equation of state can be representative of catastrophically disrupted comets.
47

Cometary Impact Lightcurves and Optical Depth Variations of
the Impact Plumes
Hitoshi Hasegawa, Satoru Takeuchi, Takuya Yamashita, Kazu Sekiguchi,
and Jun­ichi Watanabe
During the impacts of comet P/Shoemaker­Levy 9, we obtained several light curves of the
impacts at near­infrared wavelenghts at the Okayama Astrophysical Observatory (OAO) and
the South Africa Astrophysical Observatory (SAAO). Impacts of larger fragments showed
multiple #ashes. The largest impact that we observed was the fragment K, and it showed
triple #ashes. The ®rst faint #ash and subsequent secondally #ash are thought to be an entry
#ash of bolide phase and its expanding plume over Jovian horizon respectively. The ®rst
entry #ash might be missed in most of the smaller fragments. The all of the third #ash
appeared 6­7 minutes after their impact times and are the brightest. We tried to investigate
the lightcurves in terms of optical depth variation of the impact plume. A nucleation theory
in the solar composition gas was applied to the expanding plume gas. We suggest that the
secondary #ash was mainly due to gas phase opacity variation and that grain formation of
cometary originated silicate dusts will contribute a large amount of the brightest third #ash.
48

Re#ectivities of the Cometary Impact Sites
Hitoshi Hasegawa, Satoru Takeuchi, Takuya Yamashita, and Jun­ichi Watanabe
Observation of the cometary impact sites of comet P/Sheomaker­Levy 9 at near­infrared
wavelengths were carried out with a near infrared imaging system OASIS at the Okayama
Astrophysical Observatory. The observed wavelengths are the strong methane absorption
band at 2.3 micron, intermediate methane absorption at 1.7 micron, and 2.2 micron. Theese
impact sites showed as brighter clouds on the disk. The re#ectivities of several impact sites
were measured from these images at these different wavelengths. Center to Limb variation
curves indicate that their cloud top located at the stratosphere about 1 ­ 2 mbar. The material
of the aerosol particle is still unknown, but we suggest that magesium silicate dust nucleated
in the explosion plume for the most plausible candidates of the grain.
49

Near Infrared Spectroscopy and Long­Term Monitoring of the SL­9 Impacts
T. M.Herbst (Max­Planck­Institut f #
ur Astronomie, Heidelberg)
D. P.Hamilton (Max­Planck­Institut f #
ur Kernphysik, Heidelberg)
H. B# ohnhardt (Universit# ats±Sternewarte, M# unchen)
J. L. Ortiz­Moreno Instituto de Astro®sica de Andalucia, Granada
We present K band spectra taken during the H and L impacts and attempt to derive physical
conditions in the emitting regions. Prominent spectral features due to CO and methane
indicate temperatures in excess of 2000 K near maximum light. These spectral lines faded
within minutes. Monitoring of the impact sites in the days, weeks, and months after the
collision shows longitudinal spreading of the ejecta structures at a rate of 10 m/s. The
resulting band displayed signi®cant inhomogeneities as late as February 1995. Although still
prominent at 2.3 microns, the band has faded considerably since solar conjunction.
50

The Collision of Comet P/Shoemaker­Levy 9 and Jupiter
John J. Hillman, David A. Glenar, William C. Maguire, Gordon Chin,
William E. Blass 1 and Miska Le Louarn 2
Laboratory for Extraterrestrial Physics
NASA/Goddard Space Flight Center
Greenbelt, MD
1 Department of Physics and Astronomy
Univ. Tennessee, Knoxville, TN
2 Universities Space Research Association, Seabrook, MD
We observed the southern hemisphere of Jupiter on two nights during this impact sequence,
using the Coude focus at the Kirtland Air Force Base, 1.5­ meter adaptive optics telescope.
Our objectives were to recover a set of scattering properties and likely vertical distributions
of material at the impact site locations. Multispectral images were acquired in the methane
bands at 725­ and 890­nm and in their adjacent continuum using an acousto­ optic tunable
®lter (AOTF) imaging spectrometer with a silicon CCD at the focal plane. The stratosphere
near 1­mbar was simultaneously sounded by imaging in the deep 2.35­micron methane band
using a NICMOS­3 camera with cooled narrow band (0.05­micron FWHM) ®lter.
All observations were photometrically calibrated and reduced to absolute re#ectivity (I/F)
using close coincidence standard stars and published solar intensities. We subsequently
narrowed the observed point spread function in the images by employing a modi®ed inverse
Fourier deconvolution algorithm and then renormalized to the original I/F scale.
For the "quiet" atmosphere we have adopted a vertically inhomogeneous atmos­ pheric model
using two clouds and a high­level thin haze which was optimized to ®t our observed Jupiter
center­to­limb I/F variations. We then inter­ leave a collision site layer with a variable top
and bottom boundary, optical density, single­scattering albedo and phase function. Results
on several collision sites will be presented.
51

The Search for Historical Impact Sites on Jupiter
Thomas Hockey, University of Northern Iowa
A search was made through historical reports of major dark spots on Jupiter, recorded during
the pre­photographic era. While many of these reports meet the suf®cient condition for
impact candidacy (exogenic origin cannot be ruled out), the descriptions and drawings lack
the detail required to meet the necessary condition (a morphology similar to that created by
the Shoemaker­Levy 9 impacts). The lack of available information about speci®c historical
spots on Jupiter is a consequence of the perceived physical nature of these spots at the time,
and the traditional purpose of jovian spot observations.
52

Direct Observations of the Comet Shoemaker­Levy 9 Fragment G Impact by Galileo
UVS
C. W. Hord (1), W. R. Pryor (1), W. K. Tobiska (2), A. I. F. Stewart (1), K. E. Simmons (1),
J. J. Gebben (1), C. A. Barth (1), W. E. McClintock (1), L. W. Esposito (1),
R. A. West (2), S. J. Edberg (2), J. M. Ajello (2),
and K. L. Naviaux (2)
(1) Laboratory for Atmospheric and Space Physics
University of Colorado
Boulder, CO 80303
(2) Jet Propulsion Laboratory
California Institute of Technology
The Galileo Ultraviolet Spectrometer (UVS) team has detected the Shoemaker­Levy 9 frag­
ment G impact on Jupiter in data recently played back from the spacecraft tape recorder.
A 20% brightening of the disc­integrated signal of Jupiter was detected at 292 nm during a
swath across Jupiter that lasted 1:6 sec and was centered at 1994­July 18 (day 199)/07:33:31
UT (all times in this paper are corrected to be the time of the event as seen from Earth).
The emission brightness, when combined with simultaneous Photopolarimeter Radiometer
(PPR) measurements at 945 nm, is consistent with thermal radiation at a temperature of 7800
(+500, \Gamma600) K emitted over an area of 40 (+60, \Gamma25) km 2 . No excess signal was seen
during swaths 5:33 sec before and after the detection swath.
53

On Particle Acceleration by the Impact­driven Field­aligned Current System
W. H. Ip
Max­Planck­Institut fuer Aeronomie
D­37191 Katlenburg­Lindau, FRG
Following the scenario that a localized atmospheric wind system will be generated at the
comet impact sites as a result of the energy release from atmospheric explosion (Hill and
Dessler, 1995; Ip, 1995), a simple model is constructed to estimate the possible ranges of
the electric current #ows and the associated potential drop. The particle acceleration effect
caused by a potential drop of 100 eV ­ keV along the magnetic ®eld line could in turn
produce the ultraviolet emissions observed by HST after the K impact. Such ionospheric
dynamo mechanism might have some interesting implication on the energy budget of the
Jovian thermosphere as well.
54

On the Ballistic Nature of the Shoemaker­Levy 9 Impact Plumes A,E, and G
Kandis Lea Jessup (University of Michigan, AOSS)
John T. Clarke (University of Michigan, SPRL)
Heidi B. Hammel (M.I.T.)
We present the time evolution of the plumes associated with SL9 impacts A, E, and G. Our
results are based on the analysis of real time HST WF/PC2 images for each event, as well as
those of the ejecta pattern imaged 1­2 hours subsequent to each event. From these images
we have determined altitude above the limb and displacement in the plane parallel to the
"surface" of the planet for each plume as a function of time. We have also cal­ culated the
ballistic trajectories necessary to reproduce each ejecta pattern, including coriolis de#ection,
for comparison. Thus we have reconstructed the #ight of each plume, constraining the
location at which the material descends through the atmosphere of Jupiter and the initial
timing of each explosion event.
55

Comet Capture Statistics
D. M. Kary
UCSB
L. Dones
San Jose St.
Before being observed, Comet P/Shoemaker­Levy 9 had orbited Jupiter for several decades
(L. Benner, W.B. McKinnon 1994, BAAS 26, 1564). We have investigated comet capture
statistics through numerical simulations of test particles with orbits like those of Jupiter­
family comets. We did 4­body point­mass integrations (Jupiter, Saturn, Sun, and comet),
following the orbits of ¸ 3 \Theta 10 4 bodies for ¸ 10 5 years. Over 2:7 \Theta 10 5 captures occured.
Of the 2:1 \Theta 10 4 captures which make 1 or more orbits around Jupiter, ! 0:2% make 20+
orbits (corresponding to 50+ years) around Jupiter. Of the captures which get closer than
2:4R J
, 4.4% made 20+ orbits. Provided close passage is necessary for making a captured
comet observable, the bias in favor of long­term capture events is signi®cant.
This work is supported by NASA Planetary Geology and Geophysics Grant # NAGW­2061
at UCSB and by RTOP 151­01­60­09 at NASA Ames.
56

Infrared Spectroscopy of Jupiter's Atmosphere after the A, D, and E Impacts of
Comet P/Shoemaker­Levy 9
Sang J. Kim
Department of Astronomy and Space Science
Kyunghee University
Yongin, Kyunggido, Korea
and
Department of Astronomy
University of Maryland
College Park, MD 20742, USA
Glenn S. Orton
Jet Propulsion Laboratory
MS 169­237
2800 Oak Grove Dr.
Pasadena, CA 91109, USA
Christophe Dumas
Institute for Astronomy
University of Hawaii
2680 Woodlawn Dr.
Honolulu, HI 96822, USA
Infrared spectra of Jupiter's atmosphere were obtained with the Infrared Spectrometer (IRS)
on the 1.5­m telescope at the Cerro Tololo Inter­American Observatory (CTIO) during the
®rst two days of the impacts of the fragments of Comet Shoemaker­Levy 9 (1993e). We
monitored 2 ­ 4 micron radiation from the impact areas, undisturbed areas, and auroral
regions of Jupiter after the A, D, and E impacts. The strong emission of the 3 micron
band of methane was detected on the A impact area 4 hours after the impact. Emissions of
trihydrogen ions decreased at the A and E impact sites compared with undisturbed areas at
the same latitude. The temperatures of trihydrogen ions in the southern auroral region were
normal within the ®rst several hours following the A, D, and E impacts.
57

Sizes of the Comet SL9 Secondary Nuclei A and Q2 Calculated on the Basis of
Physical Parameters of Light Echoes from Europa and Io
Kleshchonok V.V., Churyumov K.I.
Astronomical Observatory of Kiev University, Ukraine
On the basis of high­speed photoelectric observations of Europa and Io, ®reball #ashes in
the atmosphere of Jupiter during the impacts of fragments A and Q2 of comet SL­9 were
recorded. The #ash of the A fragment on July 16, with an amplitude of 0.12 mag and a du­
ration of 0.7 sec, was recorded during observations of Europa. The #ash of the Q2 fragment
on July 20, with an amplitude of 0.11 mag and a duration of 1.0 sec, was recorded during
observations of Io. Analogous parameters of the second #ash were obtained at the Vatican
Observatory [1]. The data allowed us to estimate the energy of the #ashes and fragment
radii. Taking into account the paper by Sekanina [2] that in light energy of ®reball radiation
transforms 1nucleus of comet SL­9 during its impact with the Jupiter atmosphere we will
obtain the following estimate for sizes of secondary nuclei A and Q2 of SL­9: R(A)=1.42
km for ae=0.3 g/cm3 (1.00 km for ae=1.0 g/cm3); R(Q2)=0.65 km for ae=0.3 g/cm3 (0.43 km
for ae=1.0 g/cm3 ). In the paper by Hammel H.B. and Nelson R.M. [3] parameters of #ashes
of brightness of Io that had the amplitude about 0.5 mag and that was observed July 26, 1983
are given. This observation was obtained with the 1.52­m telescope through 420 nm ®lter at
the Palomar observatory.Though the #ash on Io that was registered by Hammel and Nelson
is noticeably greater then the one we registered in 1994 on Jupiter after July 26, 1983 there
was not observed a somewhat visible new spot that can be compared with those spots that
were formed on Jupiter after the collision of comet SL­9 with Jupiter. This fact is evidence
of that that the #ash of brightness of Io in 1983 was not caused by light echo from a possible
®reball on Jupiter. The most probable reason for this #ash to occur could be fall down on Io
of 1­2 km asteroid or icy cometary nucleus. As a result of this collision from Io's surface
a chain of bodies containing matter from the surface layers of Io whose composition as is
known comprise Na, S2 and other elements could be exploded. Bodies exploded from the
surface of Io, we think, could form a cometary train that in 1993 was detected by Shoemaker
and Levy as a new comet consisting of 21 secondary nuclei.
References:
1. G.J.Consolmagno and G.Menard. A Search for Light Echoes of A, H, and Q Events.
European SL­9/Jupiter Workshop. February 13­15, 1995. Garcbing, Germany, p. 25­26.
2. Sekanina Z. Disintegration phenomena expected during collision of comet SL­9 with
Jupiter. Science, 262, p.382­387 (1993).
3. Hammel H.B. and Nelson R.M. Bright #ash on Jupiter in 1983. Nature, 1, N11, p.46
(1993).
58

Near­Infrared Spectroscopy of the R Impact Site of Comet Shoemaker­Levy 9
R.F. Knacke 1 , S.B. Fajardo­Acosta 1 , T.R. Geballe 2 , & K.S. Noll 3
1 Penn State Erie, The Behrend College, Erie, PA 16563
2 Joint Astronomy Center, University Park, Hilo, HI 96720
3 Space Telescope Science Institute, Baltimore, MD 21218
We present 2.20±2.41 ¯m spectroscopy of the R event of Comet Shoemaker­Levy 9, spanning
approximately 45 minutes prior, during, and after the impact. We model the spectra with
temporally variable column abundances of mainly CH 4 and CO. The CO 2±0 bandhead is
visible during and shortly after the impact event, implying temperatures – 2000 K. The
3±1 and 4±2 CO bandheads are possibly present in some spectra, blended with CH 4 lines.
Therefore some of the CO could have reached temperatures – 3000 K. Absorption on the
red side of the 2±0 bandhead by cooler (T ú 1000 K) CO is also evident.
59

CCD Spectroscopy of Jupiter's Crash Latitude
in 4600­10,240 #
A Wavelength Region
P.P. Korsun 1 , Yu.V. Sizonenko 1 ,
S.G. Sergeev 2 , and S.V. Berdyugina 2
1 Main Astronomical Observatory, Kyiv, Ukraine
2 Astrophysical Observatory, Nauchny, Crimea, Ukraine
The impact sites of the fragments of the Comet Shoemaker­Levy 9 in the Jovian atmosphere
were observed using the 2.6­m Shajn Telescope of the Crimean Astrophysical Observatory.
We obtained long­slit CCD spectra of the spots DGRS, FTVE, H, K, KUW, L, and N at
the Nasmyth and coude foci. The Nasmyth spectra covered the spectral region from 4600
to 10,240 #
A at a resolution of 4.5 #
A. The coude spectra were obtained at the methane band
(8900 #
A) and at 5893 #
A (Na doublet) with resolution 0.85­1.7 #
A.
We investigated both the spatial structure of the impact sites along the crash latitude, and
spectral peculiarities along dispersion. All the spots were seen in absorption in the observed
spectral region except for the CH 4 bands at 8900 #
A and 10,000 #
A, where they were brighter
then undisturbed Jovian surface. Some spots showed spectral peculiarities at 7200 #
A (CH 4 )
and 7900 #
A (NH 3 ) as well. We had not detected any Na variation in our coude spectra.
60

POSSIBLE 29MHz PRECURSORS ASSOCIATED WITH SOME COMET
SL9/JUPITER IMPACTS FROM RATAN­600 OBSERVATIONS
Korzhavin A.N., Bogod V.M., Dikij V.N., Dikij D.V., Komar, N.P.
Special Astrophysical Observatory of the Russian Acad. Sci.,
Pulkovo Branch, SAO, Pulkovo, St.Petersburg, 196140, Russia
Decametric radio observations of Jupiter were made during July 16 to 22, 1994 from the
North Caucasus with the RATAN­600 radio telescope of Special Astrophysical Observatory
of the Russian Academy of Sciences. A special feeder sistem was designed with a purpose to
form a beam (with the main mirror) of about 15 degree width in the E­W direction allowing
to make an observational session of up to one hour duration. Both right­hand and left­hand
polarizations were recorded at frequency 30 MHz with 1 MHz bandwidth and 10 milliseconds
time resolution. A total of 20 observational sessions of 30 ­ 60 minutes duration have been
made. Four of them encompass impact times of some comet SL9 fragments with Jupiter and
are discussed in this paper.
Summarizing our data, we conclude that there is a quite high probability that each comet
SL9 fragment produced a precursor burst event at decametric wavelengths when passing
throughout low levels of the Jovian magnetosphere. There are also some post impact burst
events. If so, the unique event, the collision of comet P/Shoemaker­Levy 9 and Jupiter,
gives us an unprecedented opportunity to study the structure of the low levels of the Jovian
magnetosphere (the last 20,000 Km of each fragment orbit) with a spatial resolution as high
as 1­2 Km, which corresponds to our 20­40 millisecond time resolution.
61

On the spatial distribution of fragments of Comet Shoemaker­Levy near Jupiter
N.Ya. Kotsarenko, K.I.Churyumov, V.N. Mal'nev, and A.N. Kotsarenko
Kiev University, Ukraine
The fragment parameters of comet Shoemaker­Levy and their spatial distribution before
entering the atmosphere of Jupiter have been determined with suf®cient accuracy. Special
attention must be given to the geometry of the string­of­pearls chain of fragments that were
arranged along straight line on approximately equal distances between large fragments.
Evidently such spatial distribution of fragments could be a result of explosion of the comet or
could be formed by fragment interaction. Moving in strong nonhomogeneous gravitational
®eld of Jupiter fragments weakly attract and repulse each other because of elasticity of
gas atmospheres that surround each fragment. From the balance of the average attraction
force of proper gravitation and the repulsion force of the atmospheres in the stationary
regime we obtained the correlation between an average distance l i
between fragments, their
masses m i
and average density number n 0 of the gas atmosphere particles and their effective
temperatures.
It happened to be that l i
¸ p
m i m i\Sigma1
( m i
are masses of neighbour fragments). This formula
may explain the observable fact that distances between light fragments are smaller than
distances between bigger fragments. The time of the interaction transmission in this system
is estimated. More detailed comparison of the theory with the observed results may be made
if more exact density number of the gas atmosphere n 0 is known.
62

VLA Observations of the Effects of the Shoemaker­Levy 9 Impact on the Synchrotron
Emission from the Jovian Magnetosphere
A. Kundu, A. W. Grossman, J. C. L. Wang & S.M. White (Univ. of Maryland)
D. O. Muhleman and M. A. Gurwell (Caltech)
We observed the Jovian system with the VLA in the X­Band (3 cms) and the C­Band (6
cms) during the week of the impact. High resolution interferometric images of the linearly
polarized emission from Jupiter's magnetosphere were obtained. The source of the polarized
emission at these wavelengths is synchrotron emission by high energy electrons trapped in
the inner magnetosphere. The intensity and morphology of the synchrotron radiation is
dependent on the con®guration of the Jovian magnetic ®eld with respect to earth, the energy
of the electrons, and their pitch angle relative to the magnetic ®eld.
Our maps resolve the detailed structure of the emission out to 2.5 jovian radii with a spatial
resolution of 1.5 arcsecs. Snapshot images at different central meridian longitudes show
a prominent region of extended emission along the magnetic equator on either side of the
disk of Jupiter. Regions of diffuse emission are also observed at high magnetic latitudes,
due to the relativistic electrons with small pitch angle at the equator. The #ux from each of
these regions was observed to increase by a factor of 1.2­2 in the post­impact maps. The
position of the peaks of the emission migrated closer to the disk of Jupiter by up to 6 arcsecs.
Comparison of the #uxes at the two wavelengths reveal a decrease in the spectral index of
electrons with an energy of 25­35 MeV.
The synchrotron emission is observed over a more extended region, both radially and (mag­
netic) latitudinally in the post­collision images of Jupiter's magnetosphere. This suggests an
infusion of high energy electrons at large Jovian radii and possibly an inward radial diffu­
sion of the radiating electrons. The latidudinal expansion of the emission regions may by
explained by the infusion of electrons with intermediate pitch angles.
The morphology of the synchrotron emission shows strong time dependent changes cor­
responding to the changing geometry of the magnetic ®eld as seen from the earth. The
interpretation of the change in the various components of the synchrotron radiation at differ­
ent longitudes is the subject of an ongoing study.
63

10 microns Observations of SL­9 Impacts with CAMIRAS at NOT
P.O. Lagage, Ph. Galdemard, R. Jouan, P. Masse, E. Pantin, M. Sauvage (CEA,
DSM/DAPNIA/Service d'Astrophysique, CE Saclay, F­91191 Gif­sur­Yvette; e­ mail
Lagage@sapvxa.saclay.cea.fr); B. Mosser (IAP); G. Olofsson, M. Huldtgren (Stockholm
Observatory); J.A. Belmonte, C. Regulo, T. Roca Cortes, J.M. Rodriguez Espinosa, M.
Selby, L. Vidal (IAC); D. Gautier (Observatoire de Meudon); A. Ulla (NORDICA)
From July 16th to July 27th, Jupiter was imaged with CAMIRAS, the Saclay mid­IR camera,
mounted on the 2.5 m Nordic Optical Telescope (LaPalma island, Spain). The 10­13 micron
®lter and the 0.9 arcsec PFoV (total ®eld: 57x57 arcsec**2) were used. The weather
conditions were ®ne. Out of the 10 impacts observed (A, E, F, H, L, P2, Q1, Q2, T, U), 5
were detected (A, E, H, L, Q1).
The light curves obtained within 1 hour after the detected impacts are quite similar. They
will be discussed in terms of plume rising above the limb (precursor #ash about 2 min after
Galileo detection), fall­back of material into the atmosphere (main IR peak about 13 min after
impact), and sun­heated dust (bump about 20 min after impact and subsequent emission).
The prime aim of the observations was the search for thermal #uctuations associated with
seismic waves excited by the comet fragments. Direct information about the internal structure
of Jupiter can be derived from the arrival times at various distances from the impact. The
results of the search will be presented.
64

REDUCED FIVE MICRON LIGHT CURVE OF THE COMET SL­9 JUPITER
R­IMPACT FEATURE FROM THE AIR FORCE MAUI OPTICAL STATION
J. V. Lambert, V. L. Porter
Rockwell Space Operations Company
J. L. Africano, D. L. Nishimoto, R. A. Nolan
Rockwell Power Systems
P. Kervin, R. Medrano
Phillips Laboratory
A 4.8 to 5.08 micron infrared light curve of the Comet Shoemaker­Levy 9 fragment R
impact on Jupiter was derived from infrared imagery obtained using the Enhanced Longwave
Spectrometer/Imager (ELSI) on the Phillips Laboratory 1.6­meter telescope at the Air Force
Maui Optical Station (AMOS). The impact site became visible as it reached the planet's
limb, eight minutes after the estimated impact time, then faded rapidly, within ten minutes.
There is an indication of a precursor event at four and one half minutes after impact. The
two­second temproal resolution light curve contains multiple maxima. The imagery was
processed to obtain estimates of the impact feature peak intensity and diameter. Some ®ve
micron imagery of the C and F impacts was also obtained, as well as a few frames of the R
impact in the 8 to 13 micron band.
65

Jovian Decametric Radio L­bursts Associated with the Impact of Comet SL9
and Jupiter
Xiao­Cong Li and Xi­Zhen Zhang
Beijing Astronomical Observatory
Chinese Academy of Sciences
Beijing 100080
The characters of two decametric radio L­bursts possibly associated with the impact of Comet
SL­9 and the Jupiter are brie#y described. Two L­bursts occurred on July 21,1994. They took
place 55 minutes before and 13 minutes after the impact of the fragment S with the Jupiter
respectively. These two radio L­bursts have been recorded at frequencies from 24.0MHz
to 28.5 MHz by Beijing Astronomical Observatory and China Research Institute of Radio
Wave Propagation
The bursts have distinct peak #ux densities and durations. except the effect of emission
from Io satellite, The peak #ux densities are generally 7­8 times than that of a normal one,
the longest duration is 266s. The burst occurred after the impact is coresponding with the
brightening at infrared wavelength. Their spectral characters are different from that of a
normal L­burst.
The time pro®les of the two bursts have opposite frequency drifts, and the pro®les after low
pass ®ltering(0.01Hz) have opposite arc curvatures(one is early and another later). The drift
rate of the burst occurred before the impact is ­377KH z
, that of another one is +463.9KHz.
However the pro®les of two bursts are similar at ten frequencies. Their #ux densities reached
the maximum at 25 MHz. Along with the directions of frequency drift, their line pro®les of
curves became more simple, and their durations became shorter.
The spectral characters of these two bursts mentioned above seem to show that the two bursts
excited in a similar way by the impact of the fragment S with the Jupiter.
This paper was supported by The National Nature Science Foundation of China and the
Chinese Academy of Sciences.
66

Three strong Jovian decametric radio bursts from
impacts
Qi­Bin Li, Xi­Zhen Zhang, Jin­Lin Han
Beijing Astronomical Observatory, Chinese Academy of Sciences, Beijing 100080, P.R. China
April 25, 1995
Abstract
Three strong decametric radio bursts (?20db), undoubtedly related to the impacts
of fragments G, K, and W of Comet Shoemaker­Levy 1993e (SL­9) on the Jupiter,
have been detected during the impact week at Xin­Xiang temporary Jovian Decametric
Watch Station of Beijing Astronomical Observatory. All of them are narrow band
events occurred at about 26.0 and 28.5MHz, and have a long duration of at least several
minutes. Their completely different characteristics, (ie. strongest one before G impact,
immediate response to K impact, and switching bursts after W impact), are shown in
this report.
Key words: Jupiter, P/comet/SL9
67

Temporal Evolution of the SL­9 Impact Features on Jupiter from CCD Imaging:
A Video Animation
Sanjay S. Limaye & M. Lindgren
Jupiter was imaged from the solar telescope of the Swedish Royal Academy of Sciences at
La Palma (Canary Islands) during the SL­9 impact week using a broadband ®lter several
times a minute and exposures of 0.5 ±2 s. Typically more than four hundred images were
acquired each night for about four and a half hour period on ten nights. With an image
diameter of almost 600 pixels (0.06 arcsec per pixel), the scale of the spatial features resolved
on the disk of Jupiter is about 700­1400 km, limited mostly by atmospheric seeing. These
images are among the best CCD images of Jupiter obtained during the impact period from
the earth. By mapping these images into sections of global latitude­longitude maps, the
temporal evolution of the impact features is revealed dramatically, particularly when viewed
as an animated sequence. Typically, all the spots observed on their initial appearance on the
disk predominantly display an apparent drift in longitude as they move from the morning
terminator to the bright limb. The magnitude and near constancy of the drift rate on successive
rotations and the central meridian transit time measurements suggest that the actual drift rate
is small and that parallax due to immense vertical displacement of the features is a dominant
cause of the observed apparent longitude drift. By also removing the photometric function,
some additional insights into the evolution of the impact features be obtained. For example,
the magnitude of the minimum brightness of the core region appears to decrease with time
for as many as 12 to 15 hours after the impact and then slowly rise. This might be indicative
of the increasing optical thickness of the core region as viewed from the earth and then a
gradual dispersal in the vertical and horizontal directions. The video animation sequence of
mapped images illustrates the expanding ejecta around most spots and the distortion in the
apparent shape of the plume as the impact features move across the disk, most likely due to
uncompensated parallax.
68

Modeling of UBVRIJK 0 Observations of P/Shoemaker­Levy 9: Implications for the
Dust Size Distribution, Emission History, and Icy Composition
C.M. Lisse (HSTX/NASA­GSFC), M.F. A'Hearn (UMd), P.A. Esterle (UMd),
L.A. McFadden (UMd/UCSD), H.A. Weaver (STScI), L.M. Woodney (UMd)
Periodic comet P/Shoemaker­Levy9 was observed in the UBVRIJK' passbands from January
1994 through its July 1994 impact with Jupiter from the Kitt Peak National Observatory
(KPNO), the Cierro Tololo Interamerican Observatory (CTIO), the NASA/Infrared Telescope
Facility (IRTF), and the Hubble Space Telescope (HST). Extended comae and dust tails many
arc­seconds in extent were detected for some 10 of the cometary nuclei.
We have taken the UBVRIJK' images and attempted to ®t them using modi®ed Finson­
Probstein/Monte Carlo dynamical models. These models calculate the classical orbit of a
dust particle experiencing the gravitational in#uence of the Sun and Jupiter and the radiation
pressure of the Sun. The observed dust tail is then ®t by adjusting the rate of dust emission vs
time, the velocity of dust emission vs particle size (b), and the particle size (b) distribution.
The models are typically poorly constrained unless observations over a large angular scale
and a large time range (> 1 month) are made. We present conclusions concerning the emission
history and dust particle size distribution of P/Shoemaker­Levy from the dynamical models.
We have also combined the images in different passbands to create spectra of the observed
dust. We have then attempted to ®t the spectra using a Mie scattering code modi®ed to allow
for porous spheres, multi­component dust, and various particle size distributions. Cross­
checks to the particle size distributions found by the dynamical modeling were made. We
present conclusions concerning the silicate:carbon:ice composition of the observed dust in
the different cometary nuclei.
69

Five­Color Thermal­Infrared Spectrophotometry of Selected SL9 Impact Sites
T.A. Livengood 1;2 , H.U. K# au# 3 , T. Kostiuk 2 ,
G.L. Bjoraker 2 , P.N. Romani 2 , B. Mosser 4 , M. Sauvage 5
1 University of Maryland, College Park, MD
2 NASA/Goddard Space Flight Center, Laboratory for Extraterrestrial
Physics, Greenbelt, MD
3 European Southern Observatory, Garching bei M# unchen, Germany
4 Institut d'Astrophysique de Paris, Paris, France
5 SAP CEA, Saclay, France
Thermal­infrared imaging of Jupiter was conducted at the European Southern Observatory
in the ®rst 3 days of the SL9 impact week, and for a week after the impacts. Five of the
six ®lters used have proven to result in useful spectrophotometric information on the impact
sites, spanning the range 7.7±13.3 ¯m. The spectral distribution of impact­site emission
is inconsistent with blackbody thermal emission. The spectra of different impact sites are
also distinct, despite the presumably homogeneous atmospheric and impactor composition.
Presumably, the only variables between impacts were the fragment size and mass distribution
(fragment swarms versus single fragments). We will report on several impact sites and
investigate the nature of differences between them.
70

Some Results from the Ukrainian Observations of the SL9 Collision with Jupiter
D.F. Lupishko
Astronomical Observatory of Kharkiv University
Ukraine
Within the scope of the Ukraine Program the comprehensive observations of Jupiter and its
sattelites were carried out. Hundreds of distributions of brightness and polarization along the
Jovian central meridian were obtained before, during and after the collision. Two splashes of
Io brightness were recorded as a result of collision of comet fragments with Jupiter. These
data were used to estimate an explosion energy, mass and size of fragments. The fast spectral
changes (from 10 sec to 10 min) of intensity of Na, Li, Ca lines and of CH 4 , NH 3 and H 2
were detected. The decrease of methane absorption at 0.89 and 1 ¯m occured to be the overal
peculiarity of the impact sites on Jupiter. It is varied from 10­15% (medium­sized spots) to
20­50% (giant spots). The Doppler shift of some lines is evidence of very rapid movements
of atmospheric matter.
71

Thermochemical and Kinetic Modeling of the SL9 Impact Debris
James R. Lyons
Caltech
Aerosols and dust particles in a variety of environments typically have imaginary refractive
index k 1E­3 to 1E­2. The particles are (or are thought to be) heterogeneous, consisting
primarily of a weak absorber (eg. silicates,sulfates) and a small amount of a strong absorber,
usually carbon soot. West et al. (1995) reported k 6E­3 to 3E­2 for the SL9 impact
debris observed at wavelengths from 950 to 250 nm. The relatively #at k through the visible
is strongly suggestive of the presence of a soot­like compound, either as small ( 10 nm)
soot/graphite particles distributed throughtout the debris particle, or as some other species
with high carbon content. The stronger absorption exhibited by the debris particles in the
blue and near­uv is qualitatively well­matched by poly­HCN (West et al,1995).
Thermochemical analysis of a mixture of Jovian gas at temperatures 1500 to 2000 K and
pressures 1E­2 to 1E+2 suggests that a signi®cant fraction of carbon is present as graphite;
at higher temperatures little or no graphite is seen. However, very low quantites of PAH's
and higher polyacetylenes are predicted, suggesting that in a gas of jovian composition, the
kinetics may be too slow to form appreciable soot. Increasing the C/H ratio from 1E­3 (Jovian
mix) to C/H 1E­1 results in large quantities of PAH's and polyacetylenes at equilibrium,
thus providing rapid pathways for soot formation. Such a large C/H ratio could only result
from cometary material, and would only yield soot if C > O locally, as might be true for
cometary crust. While abundant HCN is expected, it seems unlikely that true poly­HCN
would form. However, given the lack of chemical data on HCN polymers, poly­HCN and
other less complex HCN polymers cannot be ruled out.
72

Case Study of Enhancement of HST Images of Pre­crash SL9 and Enhancement
Related Considerations
Stephen L. Mahan
Gordon Chin \Lambda
William E. Blass
Larry Senesac
James Rasnake
Department of Physics and Astronomy
The University of Tennessee
Knoxville, TN 37996­1200
\Lambda Laboratory for Extraterrestrial Physics
Goddard Space Flight Center, Greenbelt, MD 20771
Using the image restoration methods presented at the STScI Restoration of HST Images and
Spectra II 1993 Workshop (Chin, Mahan, and Blass, Proceedings p. 49), we have enhanced
HST WFPC­I and WFPC­II SL9 pre­crash images. Correlation of the enhanced images with
the observed impacts is presented. Attempts to recover morphological information from the
restored images will be discussed.
We shall also report on work in progress including an attempt to determine the point spread
function of John Hillman and David Glenar's AOTF imaging spectrometer coupled to the
Kirtland Air Force Base, 1.5 meter adaptive optics telescope (cf. abstract, this conference)
using enhanced HST WFPC­II images as #ground truth#.
73

3­ and 5­micron spectroscopy of Jupiter at the Canada­France­Hawaii telescope
during the collision of comet SL­9
J.P. Maillard (IAP, France), P. Drossart, B. B#ezard, C. de Bergh, E. Lellouch, A. Marten (Obs.
Paris, France), J. Caldwell (York Univ., Canada), J.C. Hilico (Univ. Bourgogne, France),
S.K. Atreya (Univ. Michigan, USA)
High­resolution spectra of Jupiter were obtained with the Fourier Transform Spectrometer
at the Canada­France­Hawaii telescope from July 17 to July 21, 1994. They cover different
spectral regions between 1.4 and 5 microns. Important modi®cations were observed in 3­
and 5­micron spectra recorded a short time after the impacts. The 3­micron spectra observed
just after impacts C and R, using a 2.5 arc­sec aperture centered on the collision sites, reveal
an unusual emission that we identify as CH 4 emission originating from a very small region
of the atmosphere (probably less than 100 km wide) heated to temperatures between 750 and
1500 K and located somewhere between 10 \Gamma5 and 10 \Gamma4 bar.
In a spectrum recorded around 4.7 microns 4.5 hours after impact L, we detect CO emissions
which are not usually present. We estimate that they are due to a local temperature enhance­
ment of the order of 100 K near the 2­microbar level combined with a huge enhancement in
the CO abundance. The spectrum also suggests a decrease of the temperature by at least 30
K between the 2­ and 200­microbar levels.
A more re®ned analysis of our observations will come from a detailed comparison of our
results with the CH 4 data of Dinelli et al. (B.A.A.S., vol.26, p.1582, 1994) on impact C, and
the CO data of Brooke et al. (B.A.A.S., vol.26, 1585, 1994) on impact L. Comparison with
observations at other wavelengths and different impacts will also be quite useful.
74

Status of the IRTF SL/9­R Seismic Wave Search
M. Marley and C. Walter
New Mexico State University
K. Wells, D. Hunten, A. Sprague, W. Hoffmann,
M. Sykes, and A.Dayal
University of Arizona
L. Deutsch
University of Massachusetts
G. Fazio
Smithsonian Astrophysical Observatory
J. Hora
University of Hawaii
The search continues for seismic waves launched by the R impact on approximately 200
7:8 ¯m images of Jupiter obtained with the instrument MIRAC2 on the IRTF. We have
constructed a hodogram (an image of mean pixel intensity as a function of angular distance
from the impact site and time) utilizing all of the IRTF data. The azimuthal averaging
substantially improves the experiment sensitivity over that reported earlier. Assuming a wave
excitation ef®ciency of 15%, our upper limit to the impact energy of the R event is about
now 4 \Theta 10 27 erg. Interestingly, this limit is comparable to that derived from observations
at the NOT telescope of the L impact employing an order­of­magnitude more 10 ¯m images
(Mosser et al. European SL/9 Meeting, 1995). The IRTF experiment thus demonstrates the
power of the 7:8 ¯m methane band for seismological observations of Jupiter.
75

Galileo PPR Observations of SL9 Impact Fragments G, H, L, and Q1
Terry Z. Martin & Glenn S. Orton
Jet Propulsion Laboratory
The Galileo spacecraft obtained a direct view of the SL9 impacts. High speed photometry by
the PPR of SL9 fragment impacts G, H, and L at 945 and 678 nm shows a characteristic light
curve shape. "Precursor" #ashes in earthbased telescopic data imply that the initial bolide
phase was not detected by the PPR nor by the Galileo instruments NIMS, UVS, and SSI.
Thus, detected light arises from the ®reball. The sharp change in slope of the PPR 945 nm
light curves after the steep initial 2 sec rise suggests that a new opacity source developed
then, due to formation of an atomic or molecular species.
76

Hydrogen Cyanide Polymers from the Collision of
Comet Shoemaker­Levy 9 with Jupiter
C. N. Matthews
Dept. of Chemistry
Univ. of Illinois at Chicago
Hydrogen cyanide polymers±heterogeneous solids ranging in color from yellow to orange
to brown to black±may be among the organic macromolecules most readily formed within
the solar system. The non­volatile black crust of comet Halley, for example, as well as
the orange atmosphere of Jupiter, might consist largely of such polymers synthesized from
HCN formed by photolysis of methane and ammonia. The dark brown color arising from
the impacts of comet Shoemaker­Levy 9 on Jupiter could therefore be caused mainly by the
presence of HCN polymers. Spectroscopic detection of these predicted macromolecules and
their by­products would strengthen signi®cantly the hypothesis that cyanide polymerization
is a preferred pathway for prebiotic and extraterrestrial chemistry.
77

MODELING OF THE IONOSPHERIC EFFECTS OF COMET D/SHOEMAKER
LEVY 9 IMPACTS WITH JUPITER
A. N. Maurellis and T. E. Cravens
University of Kansas
Theoretical modeling of the Jovian ionosphere with and without contaminant species as­
sociated with comet Shoemaker­Levy 9 (SL9) impacts is described. Introduction into the
thermosphere of species associated with the impacts such as ammonia, water, and hydrogen
sul®de has a dramatic effect on the ionosphere. In particular, the ionospheric electron density
can be reduced by a factor of 100 for reasonable contaminant abundances and H3+ column
densities display increases of about a factor of 10 or so. These ionospheric perturbations
should be present over areas of Jupiter much more extensive than the visible impact sites due
to strong thermospheric winds.
78

Recent Spectroscopic Observations of the Jovian Atmosphere
with the Hubble Space Telescope
M. A. McGrath (STScI), R. V. Yelle (NASA/Ames),
K. S. Noll (STScI), H. A. Weaver (STScI)
We present new spectroscopic observations of the Jovian atmosphere made in March and
April 1995 with the Hubble Space Telescope (HST) Faint Object Spectrograph (FOS). Ob­
servations made on 3 March 1995 at the impact latitude near the central meridian using the
0.86 00 aperture and covering the wavelength range 160±320nm show strong NH 3 absorption
that is signi®cantly enhanced relative to the pre­impact (14 July 1994) spectrum. There is no
evidence for the CS 2 or S 2 absorption observed in the G impact site ¸4 hours after G impact
(Noll et al., 1995). The 3 March spectrum also shows evidence for C 2 H 2 absorption, which
was the only absorber clearly evident in the pre­impact spectrum (Yelle and McGrath, 1995).
FOS observations on 7 April 1995 include one spectrum obtained near the limb covering the
wavelength range 220±320nm, and three spectra covering the wavelength range 160±230nm
which map the impact latitude in the N­S direction at ¸2 ffi latitude intervals. The limb
spectrum is very similar to pre­impact spectra, showing no evidence for the bright metallic
emissions seen near the limb close to the time of the S impact on 21 July 1994 (Noll et al.,
1995). The three spectra mapping the latitude of impact also show strong NH 3 absorption,
but show little evidence for variation of the NH 3 with latitude. Temporal evolution of the
NH 3 absoprtion in spectra obtained on 14 July, 18 July, 9 August, 23 August in 1994, and 3
March and 7 April 1995 will be presented.
References:
K. S. Noll et al., 1995. Science 267, 1307.
R. V. Yelle and M. A. McGrath, submitted to Icarus, February, 1995.
79

CASPIR Observations from Siding Spring Observatory, Australia
Peter J. McGregor
Mount Stromlo and Siding Spring Observatories
Institute of Advanced Studies
The Australian National University
Near­infrared images of the C, D, G, K, N, R, V, and W impacts were obtained with the
CASPIR instrument on the ANU 2.3 m telescope at Siding Spring Observatory, Australia.
A selection of images from the G and K impacts will be presented. Images at 2.34 ¯m
provide accurate timing of the impacts. Narrow band images in the 3±4 ¯m region slow
the development of an emission ring surrounding the impact sites, and a region of excess
northern auroral emission at a similar absolute latitude to the impacts. An image at 4.78 ¯m
taken 25 min after the K impact shows strong emission from the impact site, indicating that
the K fragment penetrated deep into the atmosphere. This emission is not present in an image
taken 45 min later.
80

THE VERTICAL STRUCTURE OF JUPITER'S CLOUD LAYER BEFORE AND
AFTER THE IMPACT BY COMET SHOEMAKER­LEVY 9
A. V. Morozhenko, A. S. Ovsak, and P. P. Korsun
Main Astronomical Observatory of National Academy of Sciences of Ukraine
Holosiiv, Kiiv 22, 252650, Ukraine
The vertical structure of Jupiter's cloud layers before the impact (the background) and on the
place of the fragment K impact for the atmospheric pressure 0.44 Ÿ p leq 8.0 bar has been
determined from the analysis of observetional data on the spectral re#ectivity in methane
absorption bands at 619, 727 and 880 nm. For the background the maximum particle
concentration (the volume scattering coef®cient is ¸ 3 \Theta 10 \Gamma5 cm \Gamma1 at the 1 bar level. The
effective radius of cloud particles is approximately 1.3 micron for p ? 1.5 bar. The altitude
pro®le of the scattering component of the effective optical depth (scattering optical depth)
forms by eddy mixing of the atmosphere. The value of vertical eddy diffusive coef®cient
is D = 0:5 \Theta 10 5 cm 2 s \Gamma1 for p ¸ 1 bar or for the scattering optical depth 1­11. Therefore
the altitude strati®cation of particles sizes arises and effective radius decreases from 1.3
micron ( p = 1 bar), to 0.45 micron (p = 0.38 bar). Thus the contradiction is explaned
between estimates of effective radius obtained earlier from observations of linear and circular
polarization of the planet's light and especially from the spectral dependence of the scattering
optical depth. After the fragment K impact, scattering coef®cient decreased almost twice for
p ¸ 1 bar and increased slightly for p ! 1 bar. The value of D increased up to 10 5 cm 2 s \Gamma1 . On
the level with p = 0.44 bar we have scattering optical depth 0.43. If sharp increase of D for p
! 0.44 bar is not assumed the appearance of the aerosol in higher layers of the stratosphere
(the pressure up to several mbars) cannot be accounted by rising cloud particles. Thus, the
hypothesis of its origin from the comet is very probable.
The research described in this publication was made possible in part by Grant N U4R000
from the International Science Foundation.
81

Jovian Photochemistry Following the SL9 Impacts
J. I. Moses (LPI), M. Allen (Caltech/JPL), and G. R. Gladstone (SWRI)
The collision of SL9 with Jupiter caused many new molecular species to be injected into the
Jovian stratosphere. We use a one­dimensional photochemical model to follow the evolution
of the impact­derived species. Sulfur photochemistry dominates the initial evolution, but the
longer­lived nitrogen and oxygen compounds also participate in the stratospheric photochem­
istry. We discuss the important photochemical processes operating on the impact­derived
species, identify the short­ and long­term reservoirs of the different elements, compare our
predictions concerning the temporal variation of the major compounds with observations,
and discuss the implications for aerosol formation and long­term observations.
82

Dynamic spectroscopy of Europa before the A­impact
H.K.Nazarchuk and L.M.Shulman
Main Astronomical Observatory of the National Academy of Sciences
Kyiv­22, 252650, Ukraine; e­mail: shulman@mao.gluk.apc.org
The spectral monitoring of Europa has been carried out on July 16, UT 19:16#20:02. The
1024­channel TV scanner at the 6­meter telescope (Special Astrophysical Observatory of
Russian Academy of Sciences) was used. The result of the observation is the record of the
numbers of the spectral channel where one­electron event took place each 32 msec. The
spectral range covered in one exposure is ¸ 1000 #
A approximately centered onto H±fi line:
4400#5300 #
A. Each spectrum has been #at®eld corrected and recalculated to the wave lengths
scale using the spectrum of a Ne+Ar+He lamp. So we have got ¸ 1000 monochromatic
(\Delta– ú 1 #
A) light curves for the above mentioned time period with a time resolution at ¸ 20 s.
This data can be considered like 128 sequential spectra as well.
There were no predicted impact during this time interval but the spectra differ from the
plain re#ected Solar radiation slightly changed by the albedo of Europa. Besides the Solar
Fraunhofer lines (the H\Gammafi is the strongest) there are many weak and strong non stationary
emission peaks. Among them a strange peak is detected at 4518 #
A (the most probably S + ).
It is localized in the UT 19:46±19:50 time interval when the intensity of the 4518 #
A emission
was enhanced by the factor ¸ 2.
It seems that H±fi emission in the pre­impact spectrum is present too. At least there is a
feature at 4861 #
A which is either H\Gammafi or something else. Other possible identi®cations are:
O + (4860.93), N + (4860.35), Fe(4859.75), Ca(4859.31). All these lines may occur in a high
temperature meteor #ush. Moreover, these lines may be present in the spectrum together as
well as the H\Gammafi. Unfortunately, our spectral resolution is not good enough for a reliable
choice. This emission arose two times: ®rstly at UT 19:28±19:30, and then simultaneously
with the 4815 #
A emission.
A group of emissions is seen in the range 4920±4965 #
A. They probably belong to atoms (Fe,
Ni, Na, Ca, Mg, Si), ions (S + , O + , C + , Ba + , N +
2 ), and molecules (CN, O 2 , OH, Na 2 , H 2 , S 2 ).
The spectrum of Europa seems to have features of a high temperature meteor spectrum. The
total contribution of the emissions to the integral brightness of Europa was rather small but
there was detectable spectral variations in some narrow spectral bands. It is probable that
two impacts of small undetectable fragments took place at UT 19:30 and UT 19:48. It seems
that a continuous meteor shower began approximately one hour before the A­impact.
83

Carbon monoxide in Jupiter's stratosphere after the impact of SL9
K. S. Noll, D. Gilmore (STScI), R. F. Knacke, M. Womack, S. Fajardo (Penn State, Behrend),
C. Grif®th (NAU), & G. Orton (JPL)
Two lines of the CO 1­0 fundamental band were observed in Jupiter's spectrum before and
after the impacts of comet SL9 with the CSHELL infrared spectrometer at the NASA Infrared
Telescope Facility on Mauna Kea. On the nights of August 1 and 2 (UT) the 30 arcsec slit
of CSHELL was placed parallel to latitude bands near the impact sites at approximately 45
degrees South and at a corresponding latitude in the Northern hemisphere. On the night of
August 1 we see no difference between CO lines at the two latitudes. On the subsequent
night, howver, we observe a distinct decrease in the central depth of the CO R5 and R7 lines.
The observations were obtained 24 hours after the previous night's, so they sample a different
range of longitudes. On August 2 our slit includes the fragment L impact site where strong
CO emission was observed immediately after impact (Brooke et al. 1994 BAAS 26, abstract
03.07). An explanation consistent with our observations is that weak emission features
persist and ®ll in the center of the CO absorption line. We will compare model spectra to
the observed CO line pro®les in an attempt to constrain the abundance of stratospheric CO
observed with this data. This is of particular importance since it appears that CO is the
primary oxygen­bearing molecule produced by the impacts (Lellouch et al. Nature 373, 592.,
Bjoraker private communication), though the low HST upper limit on CO (Noll et al. Science
267,1307.) suggests that uncertainties in the CO vertical pro®le remain.
84

Jupiter's 21cm Continuum Emission caused by the Collision of Comet
Shoemaker­Levy 9
C. A. Olano, J.C. Testori and F.R. Colomb
Instituto Argentino de Radioastronomia
C.C.5, 1894 Villa Elisa, Argentina
From July 13 to August 21 we have observed Jupiter at 1420 MHz using one of the IAR 30­m
single dish radiotelescopes. After the impact of fragment G, we detected a rapid increase of
the 21cm­continuum #ux, which reached the maximum (around 20the impact period and then
decayed within 30­60 days. The nature of this radiation is clearly synchrotron, originated
probably by a new population of relativistic electrons ( 1.E28 ­1.E29) injected into the
Jovian magnetosphere. The energy released by the explosions under the form of relativistic
electrons is of 1.E22­1.E23 erg. The decay time of the #ux density, when interpreted as due
to synchrotron losses, implies that the relativistic electrons were subject to a mean magnetic
®eld of 3­5 G. However, non­synchrotron losses can better account for the magnitude of the
losses required by the data.
85

Cloud perturbations as a result of the SL9 impacts
J.L. Ortiz \Lambda , G. Orton, K. Baines, P. Yanamandra­Fisher, J. Friedson.
Jet Propulsion Laboratory, Pasadena, CA 91106
T.H. Herbst
Max Planck Institute f#ur Astronomie, Heidelberg, Germany
D.P. Hamilton
Max Planck Institute f#ur Astronomie, Heidelberg, Germany
H. B#onhardt
Universitaets­Sternwarte, M#unchen, Germany
*also at the Instituto de Astro®sica de Andalucia, CSIC, Spain
ortiz@uli.jpl.nasa.gov ortiz@iaa.es
The normal Jovian vertical structure was affected by the formation of a very dark haze
at visible/uv wavelengths, mostly at stratospheric and upper tropospheric levels, but the
perturbations at deeper levels are still uncertain. We present the ®rst efforts to address this
topic, by analyzing some imaging and spectroscopic results from Calar Alto in the range 1
to 2.5 ¯m on several impact areas along with IRTF mid­IR images for the same impact sites.
Changes in molecular equivalent widths are investigated in the near IR as well as changes in
cloud opacity for longer wavelengths.
86

Spatial Variation and Time Dependence of the Temperature Structure of Impact Sites
Glenn Orton
Jet Propulsion Laboratory
California Institute of Technology
Joseph Spitale
California Institute of Technology
James Friedson, Padma Yanamandra­Fisher, Kevin Baines
Jet Propulsion Laboratory
California Institute of Technology
William Hoffmann, Aditya Dayal
Steward Observatory
University of Arizona
Lynne Deutsch
Five­College Astronomy Department
University of Massachusetts
Joseph Hora
Institute for Astronomy
University of Hawaii
The NASA Infrared Telescope Facility comet crash program included an investigation of
temperature structure at the impact sites. This experiment used the MIRAC2 mid­infrared
array camera to image Jupiter at several wavelegnths where thermal emission was dominated
by well mixed constituents: CH 4 at 7.8 ¯m to sense stratospheric temperatures near 8 mbar
and H 2 at 13.0, 17.8, 20.2 and 20.8 ¯m to sense tropospheric temperatures between 150 and
400 mbar. We present temperature sounding results from absolutely calibrated observations
which show the vertical variability of temperature as a function of time, concentrating
speci®cally on the region near the K impact site.
87

HST FAINT OBJECT CAMERA UV OBSERVATIONS OF THE SHOEMAKER­
LEVY 9 IMPACT SITES AND OF THEIR TEMPORAL EVOLUTION
R. Prang# e & C. Emerich (IAS/CNRS, Orsay, France)
D. Rego (IAS, Orsay, France and U. Michigan, USA)
F. Paresce (STScI, Baltimore)
Images of Jupiter have been taken in the UV with the Hubble Space Telescope Faint Object
Camera (FOC) from July 13, 1994, before the beginning of the comet fragment infalls, to
August 9, 1994, more than two weeks after the end of the event. The camera was used
with FUV ®lters centered on the H 2 bands, collisionally excited in the high latitude aurorae,
around 1500 #
A. The solar #ux re#ected by the atmosphere of Jupiter was also transmitted in
the long wavelength wing of the transmission function. When folded with a Jovian re#ected
spectrum, the FOC transmitted #ux peaks near 2000 #
A, and covers the 1700 to 4000 #
A range.
Therefore, the altitude range studied is higher than in visible images.
The core of fresh impact sites is darker and more extended than the streak seen in visible
images, and it is surrounded by an extended area (several arcsec), also larger and darker
than the visible plumes. They are very similar to features observed with WFPC2 in the
FUV, although less dark on average. This is consistent with the difference in altitude range
sampled, and can be used for a comprehensive diagnostic of the plume material.
Two images, taken on August 9, on both sides of Jupiter, show the long term evolution of
the UV absorbing material in the upper stratosphere/thermosphere. The evolution is very
different depending on the fragments, some of the plumes (as A) have almost disappeared
whereas others (the major ones) are still very dark (up to 70% absorption). In addition to
longitudinal drift of the material in the stratosphere creating a belt­like feature, we identify a
strong latitudinal diffusion with signi®cant absorption from the auroral zone up to the edge
of the FOC frame (' 20­25 c irc). Finally transient and/or meridional winds in the upper
stratosphere are inferred from the observation of several structures drifting north of the impact
region. This also is consistent with results obtained from FUV WFPC2 observations.
88

MAGNETIC MAPPING AND INTERPRETATION OF AURORAL SIGNATURES
OF COMET SL9 IN THE JOVIAN MAGNETOSPHERE
R. Prang# e (IAS, Orsay, France)
I. Engle (Naval Academy, Annapolis, USA)
M. Dunlop & M.K. Dougherty (Imperial College, London, UK)
S. Maurice (ESTEC, Nordjwick, Netherlands)
W. H. Ip (MPI, Lindau, Germany)
J.T. Clarke & G.E. Ballester (U. Michigan, USA)
Before impacting the dense atmosphere of Jupiter, the fragments of comet Shoemaker­
Levy 9 spent a few days #ying across its magnetosphere (several millions of km across).
Electrodynamic interaction with the ambient plasma induced several unique phenomena
detected in the UV, X ray and radio wavelength ranges.
Among them, the detection of an unusual 'blinking' UV bright spot in Hubble Space Telescope
images of the southern polar cap on July 20, just before P2 collision, might be attributed
to auroral type processes triggered by the charged environment of the comet fragments, as
described in some prediction models. A detailed modeling of the time varying morphology
of the magnetic ®eld lines enables us to compare the location of this spot with the magnetic
footprint of the fragments still in the magnetosphere, and to determine the characteristics of
the corresponding ®eld line. We show that fragment Q is a good candidate, and we discuss
whether the data can allow us to distinguish between the various fragments.
We establish also that Q was on an open magnetic ®eld line at the time of the observations, in
agreement with the lack of observable conjugate emission from the north, although the most
distant fragments may be on closed ®eld lines. We study the deformation of the instantaneous
Q magnetic ®eld line in the hours following the UV observations, to investigate a later set of
images and any possibility of relationship between this UV event and a close­in­time X ray
emission from the northern hemisphere. We ®nd that the ®eld line passing through Q closed
during two periods before the fragment impacted the atmosphere, the ®rst one corresponding
to the north X ray burst and to the second set of UV images.
Under the assumption that this auroral emission is indeed a signature of the interaction of
the coma of fragment Q with the magnetosphere, and that this interaction is triggered by
a mechanism alike the one predicted in Ip and Prang#e (1994) which involves ®eld­aligned
currents, we propose an explanation of the modulation of the spot brightness.
89

On the Disappearance of the Dark Band in the Jovian Southern Hemisphere before
the SL9 Impacts on the Planet
PROKOF'EVA V.V.
Crimean Astrophysical Observatory
TARASHCHUK V.P., CHURYUMOV K.I.
Astronomical Observatory of Kiev University
It is known that Jupiter disc is rich in numerous details. Some of them change their contours
appear and disappear within small time intervals. Other details, though they can suffer several
changes, remain for month, years, centuries. They all are purely atmospheric formation
because Jupiter does not have hard surface. As usual during falldown of fragments of
comet Shoemaker­Levi 9 on Jupiter dark bands and light zones between them were observed.
However blurring and later disappearance of the dark band in the southern hemisphere were
observed. This phenomena occurred before fall down of 21 secondary nuclei, i.e. it took
place before July 16 1995. After their fall down on to the visible surface of the planet
dark formation, on which the spectrograph diaphragm have been guided. The principal
components of the Jupiter atmosphere are hydrogen H2, helium He, methane CH4, in the
gaseous state and ammonium NH3 in form of aerosol of the cloudy sheet, comprising small
crystals as well as product of their dissociation as the result of photochemistry processes
(C2H2 ­ acetylene, N2H4 ­ hydrazine et al.) yield the most powerful absorption bands in
the visible range of the spectrum: 619, 725 nm ­ methane, 644.1 , 790 nm ­ ammonium).
Sharp contours of the dark features are connected with methane, while NH3 yields more
light clouds located higher and having lower temperature. Intensities of bands of methane
and ammonium is dependent not only from their abundance but to a greater extent from the
physical conditions in the atmosphere. Possible mechanism of the disappearance of the dark
band in the Jupiter south hemisphere is discussed.
90

Pre­impact Images of Comet P/Shoemaker­Levy 9 from Las Campanas
David L. Rabinowitz
Carnegie Institution of Washington, Department of Terrestrial Magnetism
5241 Broad Branch Rd, NW, Washington, DC 20015
On the night of 1994 July 16 ( 04:00 UT), visible light images of nuclei E, G, L, P2 (8a and
8b) , Q1, Q2, R, S, and W of comet P/Shoemaker­Levy 9 were obtained using a CCD and
coronagraph at 100" Dupont Telescope of Carnegie Observatories (Las Campanas, Chile).
The coronagraph was built by Steve Larson (University of Arizona) as part of the Comet
Impact Network Experiment [1, 2]. Each of nuclei E, G, L, Q(1,2), and W show trails of
material extending towards Jupiter. Similar trails appear in Hubble Space Telescope images
of the "P­Q complex" on July 20 UT, but not on July 11 UT and prior images. Because these
trails extend asymmetrically from the nuclei, it is unlikely that Jovian tidal forces can explain
their appearance. Other forces, such as Lorentz drag in the the Jovian magnetic ®eld, may
be responsible.
[1] S. Larson et al., 1994, The Comet Impact Network Experiment, BAAS 26.
[2] J. Scotti et al., 1994, Preliminary Results from The Comet Impact Network Experiment,
BAAS 26.
91

Deep Coma Imaging of Comet Shoemaker­Levy 9 to Characterize Dust Out#ow
Velocities and Grain Sizes
T. W. Rettig, J. Hahn, & G. Sobczak
Univ. Notre Dame
M. J. Mumma
NASA/Goddard
The comae resulting from the approximately 20 fragments of Comet Shoemaker­Levy 9
present a unique opportunity to characterize the ejected cometary dust grains that were once
buried deep within the parent progenitor. We present an analysis of the spatial brightness
pro®les of P/S­L 9's inner comae extracted from Hubble Space Telescope images obtained
from January through July 1994. These pro®les reveal that most of the S­L 9 fragments
exhibit the usual r \Gamma1 spatial dependence in their inner r Ÿ 1 00 comae in observations where
jovian tides are unimportant. Beyond this distance, the comae lose their circular appearance
as radiation pressure drives the dust tailward. This distance is related to two unknown
properties of comet dust grains: the grain size and out#ow velocity. But when the comet
fragments are within about 100RJ of Jupiter, where RJ is one Jupiter radius, jovian tides
become the dominant perturbation acting on the comae. However, the projected distance
beyond which tides distort the comae are related only to a single unknown, the dust out#ow
velocity. Thus, by measuring the distance beyond which the comae are initially distorted by
radiation pressure, and in later observations by tides, we are able to extract an estimate of the
characteristic dust grain size, which is of order 10 to 100 microns, and its out#ow velocity
3 meters/sec. In comparison to gas driven cometary out#ows measured in other comets, the
emitted dust grains leaving the P/S­L 9 fragments are unusually large with considerably lower
velocities. In addition to the dust sizes and velocities, the surface brightness pro®les will
be used to extract the central brightness excesses which are a measure of the re#ecting area
in the inner comae regions. We will also present the temporal magnitude variations of the
inner comae regions. These observations are being used in conjunction with tidally distorted
Monte Carlo models to place constraints on the dust out#ow velocities and dust production
rates.
92

Evolution and Drift of the Visible Impact Scars
John Rogers
British Astronomical Association
Most of the visible SL9 impact sites lasted more than a month, and their average rotation
period was very close to that of the S.S.S. Temperate Current or System III. However, there
was evidence for diverse local motions. At ®rst, the black core regions had an average System
III drift of +6 deg/month (+/­ 5, s.d.) (omitting the multiple impact sites K/W and D/G/S).
But the leading edges, peripheral clouds, and whole sites later, moved faster: ­10 deg/month
(+/­ 12, s.d.), which is faster than the underlying cloud­top currents. By mid­September the
sites had merged into an uneven new belt. In early 1995, this belt was still present but fading.
93

Normal Patterns of Dark Spots in Jupiter's Atmosphere
John Rogers and Michael Foulkes
British Astronomical Association
The SL9 impact scars were unprecedented in the experience of visual observers of Jupiter.
All available observations of Jupiter have been surveyed (J.R., The Giant Planet Jupiter, CUP,
1995). The record is complete from 1878 onwards. Almost all dark spots are of recognisable
types in ®xed latitudes moving on permanent currents. There are no outstanding records
of large black sudden­onset spots like impact sites G or L. Therefore the frequency of such
impacts (allowing for unoservability during solar conjunction) is less than one per 70 years.
94

Atomic Line Emissions in the Impacts of Comet SL­9 into Jupiter
M. Roos­Serote, A. Barucci, J. Crovisier, P. Drossart,
M. Fulchignoni, L. Lecacheux and F. Roques
Observatoire de Paris­Meudon
High­resolution spectroscopy of Jupiter was performed during and after the impacts of comet
SL­9 at the Pic du Midi 2­m telescope. Spectra with a resolving power of 36,000 between
564 and 875 nm were recorded with the MUSICOS (MUlti SIte COntinuous Spectroscopy)
spectrograph. Transient emissions of atomic lines following impacts L and Q1 were detected:
sodium (589 nm doublet), iron (multiplets at 804, 636 and 550 nm), calcium (657 nm
intersystem line and 616 nm triplet), lithium (671 nm doublet), potassium (766 nm line) as
well as the hydrogen H alpha line. These lines were detected about 15 min after the impacts
and were no longer visible one hour after.
The excitation process for these atomic emissions is not yet understood in every detail. The
resonant #uorescence mechanism can neither account for the observed intensities nor for
the presence of the Ca intersystem line. Collisional excitation is not ef®cient at the high
altitudes reached by the plumes. The falling back of the plumes at a velocity of 10 km/s
on Jupiter might explain the transient emission. Electronic recombination processes and
chemical reactions might also play a role.
The alkali and metals observed are not usually present in Jupiter's upper atmosphere and
should come from the refractory compounds of the comet nucleus. Correct modelling of the
excitation processes of the emission lines will allow the determination of the abundances of
these species in the comet.
95

DISRUPTION OF SL9 BY VOLATILES RELEASED BY TIDAL HEATING
T.V. Ruzmaikina
Lunar and Planetary Laboratory
University of Arizona, Tucson 85721
The comet Shoemaker­Levy 9, revealed in March of 1993 as a string­of­pearls of 21 major
fragments, has been destroyed when it passed Jupiter within Roche limit, at the distance
about 1.6 of the planet's radius in 1992.
In this paper we suggest an idea that the comet might had been disrupted not by the tidal
force itself, but by the pressure of volatiles released due to the tidal heating. We consider the
comet as a conglomeration of blocks divided by narrow "fracture" zones, and assume that the
tidal force initiated the sliding of parts of the comet along one (or few) zones. Then the heat
released in a narrow shear layer between the blocks could be suf®cient for the evaporation
of highly volatile elements and their pressure could disrupt the comet. The momentum
transferred to the pieces of the comet is large enough to provide them velocities exceeding
escape speed from the surface of the comet and hence prevent their coalescence, provided
that the comet radius was ! 10 km.
We speculate that the shear induced release of volatiles could also cause disruption of the
comet fragments in the upper layers of Jovian atmosphere, and thus explain the shallow
penetration of the fragments without putting a strong restriction on their sizes.
96

MORPHOLOGY AND MOTIONS OF THE SL9 IMPACT CLOUDS
(JULY 1994­APRIL 1995)
A. Sanchez­Lavega (1), J. Lecacheux (2), F. Colas (3), J. M. Gomez (4), P. Laques (5), I.
Miyazaki (6), and D.C. Parker(7)
(1) Dpto. Fisica AplicadaI, Universidad Pais Vasco, Bilbao, Spain.
(2) DESPA, Obs. Paris­Meudon, France.
(3) Bureau des Longitudes, Paris, France.
(4) Grup d'Estudis Astronomics, Barcelona, Spain.
(5) Observatoire Pic­du­Midi, Bagneres­de­Bigorre, France.
(6) Oriental Astronomical Association, Okinawa, Japan.
(7) Association Lunar Planetary Observers, Florida, USA.
A large number of CCD images of the SL9 impact clouds in the wavelength range 600­900 nm,
taken at Pic­du­Midi Observatory in France from July 1994 to April 1995, has allowed us to
study the morphological changes and motions of the impact clouds. Our measurements show
the existence of at least three different kind of motions (July­August 1994): (1) Eastward
expansion during the ®rst two days of the largest impact regions (K, L, G) with velocities
in the range from 36 to 60 m/s; (2) Clouds that acted probably as passive tracers of the
stratospheric #ow (velocities in the range ­15 to 20 m/s); (3) Local motions of the cloud
elements with meridional components, sometimes related to the interaction with anticyclonic
ovals in nearby latitudes (velocities of about 20 m/s). The average zonal velocity relative to
System III of the cores of impacts E, A, C, W, K, L, G, R, Q1 and H was (up to August 18)
­1.7 m/s with an average jovicentric latitude of ­43.8 deg. Since mid­August and up to early
October 1994, the preferentially zonal spreading of the clouds formed a nearly complete belt
at latitude ­44.6 deg together with scattered clouds alongside parallels ­55.0 and ­36.2 deg.
Post­solar conjunction observations between January and April 1995, show the permanence
of the heterogeneous dark belt at ­44 deg which contains numerous local condensations (size
3000­10000 km) dispersed along the whole latitude circle. Other synoptic clouds, probably
pertaining to the SL9 impact, were detected at ­54 deg during this period.
97

Comet Shoemaker­Levy 9 in Historical Context:
The Record of Split Comets on the Galilean Satellites
Paul M. Schenk (LPI, Houston, TX), Erik Asphaug (NASA ARC, Moffett Field, CA), William
B. McKinnon (Wash. U., St. Louis, MO), H.J. Melosh (LPL, U. of Ariz., Tucson, AZ), and
Paul Weissman (JPL, Pasadena, CA)
Shortly after discovery of comet Shoemaker­Levy 9 in 1993, Melosh and Schenk concluded
that crater chains on the Galilean satellites Callisto and Ganymede were the result of the
impact of past split comets. We investigate the morphology of crater chains, focusing on
three key problems:
ffl What are the gross physical properties (e.g., mass, radii) of Jupiter­family comets?
ffl What are the characteristics of individual fragments at the time of impact?
ffl What is the nature of the breakup process: can we distinguish between 'rubble piles,'
'solid' snowballs and other breakup models?
We compare these properties with those of Shoemaker­Levy 9, providing a historical context
in which to evaluate that event.
98

COLLISION OF COMET SHOEMAKER­LEVY 9 WITH JUPITER: IMPACT
STUDY OF TWO FRAGMENTS FROM TIMING OF PRECURSOR EVENTS
Zdenek Sekanina
Jet Propulsion Laboratory, California Institute of Technology
Pasadena, California 91109, U.S.A.
The impacts of fragments K and R of comet Shoemaker­Levy 9 are investigated with the aims
to interpret the timing of the observed precursors to the main thermal emission event and
to correlate the results of ground­based infrared observations with a variety of observations
made onboard the Galileo space­ craft. Analysis of the phenomena associated with the impact
and explosion of fragment K shows that there is no discrepancy in the timing of the Earth­
and Galileo­based observations and that the time of 53 \Sigma 3 seconds between the emission
peak of Precursor 1 and the onset of Precursor 2 can be interpreted as the interval between the
impactor's disappearance behind the Jovian limb and the ®rst appearance of the ejecta's plume
over the limb following the explosion of the impactor's residual mass, as viewed by terrestrial
observers. It is concluded that the impactor exploded at an altitude of 45 to 50 km above the
pressure level of 1 bar and that the residual mass involved in the explo­ sion, approximately
6 to 7 million tons and about 400 meters across, repre­ sented only a fraction of 1 percent of
the fragment's preatmospheric mass. The explosion is calculated to have taken place under a
dynamic pressure of several hundred bars and the dissipated energy is found to have been on
the order of 10 26 erg. The results for fragment R show it to be smaller and less massive than
fragment K, exploding slightly higher in the Jovian strato­ sphere, 50­60 km above 1 bar. The
preferred solutions suggest that the rate of ablation of these impactors was comparable with,
or somewhat higher than, that of category IIIb ®reballs in the Earth's atmosphere. These
®reballs represent a population of objects consisting of "soft" cometary material, whose bulk
density is typically 0.2 g/cm 3 . Preliminary evidence from other observations of the various
fragments appears to be consistent with the pres­ ent conclusions. All plume­expansion
models based on penetrations below the clouds are incorrect and need major revisions. The
successful prediction of explosion altitudes for the Shoemaker­Levy 9 fragments, based on
the slightly modi®ed fundamental equations of the classical theory of meteor physics and
on ablation rates derived from data on relevant terrestrial ®reballs, is a tribute to the meteor
theory and demonstrates the versatility of its tech­ niques in applications.
99

SAAO K­band Imaging Observations of the Collision of
Comet P/Shoemaker­Levy 9 and Jupiter
Kazuhiro Sekiguchi (NAO­Japan & SAAO), Ian S Glass (SAAO),
and Satoru Takeuchi (Kyusyu Univ.)
We present results of K­bandimaging observations of the collision of CometP/Shoemaker Levy 9
obtained using the SAAO/U­Tokyo/NAO­Japan PtSi IR camera on the 75­cm telescope at the
SAAO Sutherland. Time resolved (30 sec and 10 sec interval) observations of the impacts by
the fragments, A,E,H,P2,Q1,Q2,S and T were analyzed and K­band light curves of the each
impact event have been constructed.
100

Interaction between the magnetospheres of comet SL9 and Jupiter
A. S. Sharma, G. M. Milikh and A. S. Lipatov
Department of Astronomy
University of Maryland
College Park, Maryland
The plasma in the cometary magnetosphere formed by the mass loading of the solar wind by
the cometary ions is tenuous and has low energies at 5 AU. Simulations using hybrid (particle
ions and #uid electrons) plasma simulation codes have given plasma density of 2±3 cm­3 and
electron temperature of a few eV in the cometary bow shock region. The interaction of the
Jovian and cometary magnetospheres, simulated with these codes, yields local enhancements
of the magnetic ®eld and plasma density by a factor of 4 ± 5 and the electron temperature by
2± 3. Enhancement in the intensities of the ion emissions in this region was predicted at the
SL9 ­ Jupiter encounter and the HST/FOS observations have shown a strong MgII 280 nm
emission at 53 RJ from Jupiter. The detailed modeling of the emissions and its implications
to the magnetospheres at the interaction will be presented.
101

Global Maps and Drift Rates for Jovian Features
Simon, A. A., Beebe, R. F., & Chanover, N. J.
New Mexico State University
Using Hubble Space Telescope Cycle 4 data obtained on February 13, 1995 with WFPC2,
a maximum resolution global map of Jupiter has been constructed at a wavelength of 410
nm. This map can be compared to a similar map produced from images taken after the
Comet P/Shoemaker­Levy 9 impacts on July 23, 1994. Standard features have been tagged
on the February 13th map, and their drift rates computed. This is important in studying cloud
morphologies and transient features on Jupiter. In addition, interacting features have been
tagged on the impact map and their circulation rates determined.
102

Radio Observations of Jupiter from HartRAO
D.P. Smits & G.D. Nicolson
Monitoring of Jupiter's microwave emission at wavelengths of 18, 13, 6 and 3.6 cm com­
menced at the Hartebeesthoek Radio Astronomy Observatory (HartRAO) a few weeks prior
to the collision of the fragments of comet Shoemaker­Levy 9 into the planet, continued daily
during the week of impacts and irregularly thereafter. Short­timescale 6 cm and 3.6 cm data
collected during the times of 2 impacts show no observable changes. The horizon­to­horizon
runs recorded during the week of impacts show a gradual #ux increase. The largest increases
occurred at 18 and 13 cm suggesting that the synchrotron emission has increased. Monitor­
ing of the #uxes after the collisions shows that they have remained in their high state. The
analyzed data will be presented together with recent radio data (1995) from Jupiter collected
at HartRAO.
103

Near­IR Images and Spectra of the Impact Sites from Cerro Tololo
John R. Spencer (Lowell Observatory), Darren L. DePoy (Ohio State University),
Glenn S. Orton (JPL), Christophe Dumas (University of Hawaii),
Sang J. Kim (University of Maryland)
We used the 1 ­ 2.5 micron Ohio State Infrared Imager Spectrometer (OSIRIS) to observe the
Shoemaker­Levy 9 impacts from the CTIO 4­meter telescope in Chile. We obtained 1.58, 1.7,
and 2.3 micron images and 1­2.5 micron cross­dispersed R=500 spectra of the impact sites
during the following times in July (given in fractional days): 16.95±17.15; 17.93±18.15;
19.00; 22.80±22.90; and 23.18; also a few images near days 25.0 and 27.0. Continuous
movies at 1.7 or 2.3 microns during the B and F impacts showed no detectable effects due
to these events, though clouds caused brief interruptions, and all other impact events were
clouded out.
2.05 ­ 2.30 micron spectra of the impact site clouds show CH4 absorption, presumably due to
CH4 above the impact clouds: simple re#ecting­layer models assuming normal atmospheric
CH4 abundances above the impact sites give cloud altitudes in the 1 mbar range, though
there is considerable variation in the strength of the CH4 bands between impact sites of the
same age: the D site shows much stronger overlying CH4 absorption than the E site when
both are 0.48 days old, for instance. There is very little CH4 absorption above the A site
at age 0.17 days. Spectra of the E site appear to show that the #halo# around the site was
at a higher altitude than the central #core#. In contrast to visible continuum wavelengths
where the impact sites appeared dark, the sites were invisible at the near­infrared continuum
wavelength of 1.58 microns.
Follow­up imaging of the impact sites from the IRTF in September 1994, February 1995,
and March 1995 showed that considerable structure persisted in the impact­generated clouds
even eight months after the impacts.
104

Measurements of Water in the ‘Re­entry phase' Following Fragments R and W of
Comet Shoemaker Levy 9
A.L. Sprague 1 , G.L. Bjoraker 2 , D.M. Hunten 1 , F.C. Witteborn 3 , R.W.H. Kozlowski 3 , and
D.H. Wooden 3
1 LPL, The University of Arizona
2 NASA GSFC 3 NASA ARC
We have measured the abundance of water seen in spectra following the R and W impacts of
Comet Shoemaker Levy 9 into Jupiter's upper Atmosphere. Measurements were made with
HIFOGS on board the Kuiper Airborne Observatory that #ew out of Melbourne, Australia
on July 21 and 22, 1994.
The 6.6 ¯m band of high quantum number rotational states was seen for approximately 5
minutes following both impacts. Abundances are: ¸ 5.7 \Theta 10 10 g for R+12 minutes and ¸
2.2 \Theta 10 10 g for W+10 minutes. These masses are equivalent to ice balls of diameter 50 and
35 m respectively.
If we accept the Galileo NIMS conclusion that the explosion of R occurred near the 0.2 bar
level then the water is from the bolide rather than the Jovian atmosphere.
105

Polarimetric Imaging Observations of Impact Dark Spots
B. Suzuki, J. Watanabe, T. Sasaki and H. Kurihara
We carried out a polarimetric observation of traces of the comet impacts on 20th July 1994
with OOPS(Okayama Optical PolarimetrySystem) attached to the 91cm re#ector at Okayama
Astrophysical Observatory. The measurements of the degree of linear polarization fragments
K and L sites were obtained with four wavelength range(514, 644, 659 and 820nm). The
trend of maximum polarization degree of impact dark spot appears about 4 percent which
is higher than the Jovian atmosphere situated equal latitude. Wavelength dependence of
polarization is not known well from our preliminary results.
106

NUMERICAL SIMULATIONS OF IMPACT OF COMET SHOEMAKER­LEVY 9:
COMPARISON WITH OBSERVATIONAL RESULTS
T. Takata
Univ. of Tokyo
JAPAN
T.J. Ahrens & J.D. O'Keefe
Caltech
Pasadena CA 91125
G. S. Orton
JPL
Pasadena CA 9110
Numerical simulations of impact of fragments of SL9 were conducted to predict the impact
phenomena and the observational possibilities [Takata et al., 1994, Ahrens et al., 1994(1)(2)].
These results are compared with observational results, such as chemical abundance and plume
heights (Hammel et al., 1995, Carlson et al., 1995) to investigate the nature of SL9 fragments
such as the progenitor size and the origin of the parent body. The comparison suggests that
the material of SL9 is primitive, and that sizes of a large fragment and the parent body are
¸2 and ¸ 4 \Gamma 5 km in diameter, respectively.
107

The OAO near­infrared observations of the SL­9 impact to Jupiter
Satoru Takeuchi
Department of Physics
Kyushu University
Hitoshi Hasegawa
Astec, Inc.
Jun­ichi Watanabe and Takuya Yamashita
National Astronomical Observatory of Japan
We observed the impact of comet Shoemaker­Levy 9 into Jupiter by 188cm telescope of
Okayama Astrophysical Observatory with a near­infrared camera, OASIS. Flashes originated
to the impacts of fragments C,D and K were detected at wavelength of 2.35 micron. We
report the detail of these observations and discuss some impact phenomena.
108

Was SL­9 a Jupiter Family Comet or an Escaped Asteroid?
(Dynamical Considerations)
Gonzalo Tancredi and Andrea Sosa
Depto Atronomia, Fac. Ciencias
Montevideo, Uruguay
SL­9 was an unusual object in many senses. Despite the fragmentation and collision with
Jupiter, the dynamics of SL­9 were remarkable. SL­9 experienced a long­lasting temporary
satellite capture by Jupiter. The duration of the encounter can not be unequivocally determined
due to the chaotic character of the motion, but the evolution can reliably tracked for more
than 20 years. There are only two other objects which had suffered comparative but much
shorter satellite capture: P/Helin­Roman­Crockett and P/Gehrels 3. Long­lasting satellite
captures are a rare phenomena among the observed Jupiter family comets, because there is a
very narrow window in the orbital elements phase­space for which it can occur.
A useful parameter to classify the candidate orbits is the Tisserand parameter (T), an integral
of motion in the Restricted Three Body Problem. Values of T ¸ 3 are needed for long­lasting
captures. In fact, SL­9 and the two objects mentioned before have T – 3. There are also two
large populations of objects with T ¸ 3, i.e. the Trojan and the Hilda asteroids; but the mean
motion resonances with Jupiter (the 1:1 and 3:2 respectively) act as a protection mechanism
that prevents the close encounter. Nevertheless, due to mutual perturbations or collisions, an
asteroid can escape from the resonance and become a Jupiter approacher.
We compare the probability for objects in the three groups to experience a longlasting satellite
capture and estimate the frequency of collision with Jupiter.
109

THE REFLECTIVITY OF THE AEROSOL MATERIAL IN THE G, L AND E
IMPACT­REGIONS
Tejfel V.G., Kharitonova G.A., Sinyaeva N.V.,
Gajsina V.N., Aksenov A.N., Kirienko G.A.
Fessenkov Astrophysical Institute, Kazakhstan National
Academy of Sciences, Alma­Ata, 480068 , Kazakhstan
The consequences of the Comet Shoemaker­Levy and Jupiter collisions (CJC) were observed
photographically during 16 nights between 17 July and 15 August 1994 with the 1­meter
telescope in the high mountain Assy observatory (80 km eastern Alma­Ata, altitude 2750 m).
150 photographic images obtained with the scale 1.8 arcsec/mm were measured on autom­
atized microdensitometer. Computerized images were analyzed to ®nd the planetographic
coordinates of the dark impact­regions, their sizes and relative intensities in the blue and
red light. The best images (24 and 25 July) in blue and red light with the impact­regions
G, L and E placed near the central meridian were studied especially to estimate the mini­
mum relative intensities at the centers of these regions. These intensities are not less 0.62,
0.64 and 0.72 for G,L and E regions respectively on the original images and cannot be less
than O.2­0.5 if the smoothing will be considered. The modeling calculations for the most
probable characteristics of the cometary solid material that has low albedo have given the
estimates of the optical depth for the stratospheric layer formed by dark cometary material.
For assumed law of the radial distribution of the dark material amount within observed area
of the inpact­regions the sizes of the cometary fragments are derived. They cannot exceed
0.2­0.3 km and this estimate don't contradict to the values derived from other considerations.
110

High temporal resolution near­IR observations of impacts H and Q from Calar Alto
and interpretation
G.P. Tozzi, A. Richichi, & A. Ferrara
Osservatorio Astro®sico di Arcetri
Observations with a fast IR photometer at the 2.2m telescope on Calar Alto have yielded
high temporal resolution (sampling typically 100 Hz) lightcurves for some of the impacts,
including A, H, Q1/Q2, T (Richichi et al. 1995). In this contribution, we analyze in some
detail the data for the main impacts in this set, namely H and Q1. By comparing our H data
with those available at other wavelengths, we present estimates of the temperature of the
phenomenon and its evolution. In the case of Q1, it is particularly interesting to examine
the ®ne structure of the lightcurve, which is generally lost in other observations available at
lower temporal resolution.
111

FUV Spectra of SL9 Impact Site G with HST/GHRS
L. M. Trafton, S. K. Atreya, M. A. McGrath, G. R. Gladstone,
J. J. Caldwell, K. S. Noll, H. A. Weaver, R. V. Yelle,
C. Barnet, and S. Edgington
Spectra using grating G140L obtained within hours of the impact of fragment G of comet
SL9 with Jupiter are presented at effective resolution 4:4 #
A. Below 1610 #
A the spectrum
is dominated by emission, with Io's torus being tilted largely out of the ®eld of view. The
H 2 dayglow spectrum is analyzed and subtracted from Jupiter's spectrum to search for non­
hydrogenic emissions such as CO and S. Upper limits are placed on several species of interest
and the results are used to interpret impact phenomena.
112

Multiwavelength Observations of S±L 9 Impact Spots on Jupiter
R. Vasundhara 1 & Pavan Chakraborty 2
Indian Institute of Astrophysics
Bangalore 560034
1 e±mail: rvas@iiap.ernet.in
2 e±mail: pavan@iiap.ernet.in
Observations of Jupiter made at the cassegrain (f/13) focus of the 102 cm telescope at the
Vainu Bappu Observatory during 18 ± 22 July 1994 and 13 ± 14 Aug. 1994 are reported.
The images taken in July, were recorded through six narrow band ®lters centered at 4862 #
A,
4935 #
A, 5083 #
A, 6581 #
A, 8900 #
A and 8930 #
A with FWHM ranging between 50 to 300 #
A. The
differential wave length dependence of the scattered light from the impact regions compared
to nearby satellites and the surrounding regions on Jupiter are presented. Subsequent, images
obtained during 13 ±14 Aug, 1994 at 4862 #
A and 8930 #
A are compared with the July images
to look for changes in the wavelength dependence and size of the spots.
113

Comet Shoemaker­Levy 9: A New Class of Object
Siacho Wang (PMO)
The Comet SL9 has high abundance of sulfur and low abundance of H2O, CO and nitrogen,
but most of the fragments were embedded in circularly summetric inner comae from July
1993 until late June 1994, implying that there was continuous and weak outgassing activity.
This indicates that comet SL9 is a new class of object which is different from known comets
and asteroids. The main features of SL9 are 1) low abundance of H2O, CO and nitrogen;
2) high abundance of sulfur; 3) existence of central coherent body, not swarm of debris; 4)
fragile nature; 5( low albedo; 6) containing carbonaceous and silicate material; 7) low volatile
object. The existence of this new class object of SL9 indicated the variety of the objects in
the solar system.
This work was mainly supported by Chinese Academy of Sciences, Chinese National Science
Foundation and Chinese Paci®c Insurance Co. Ltd.
114

Observations and Studies of Chinese Jupiter Watch
Sichao Wang (PMO), Bochen Qian (BAO) and Keliang Huang (NUU)
China
Optical observations: 14 telescopes at 8 observatories or stations joined the CJW. Using the
CCD camera and CCD video camera, observations of dark spots have been made at SAO,
YAO, PMO, BAO and some stations during the period between June 15 and Sept. 15, 1994.
6 Satellite #ash observations have been succeeded at SAO (China) and 9 light curves of
Galilean satellites have been obtained.
Radio observations: 5 radio telescopes at 5 observatory or stations joined the CJW. During
the impact periods, a steady #ux density increase of Jupiter was seen at 12cm and 13cm
wavelength at BAS and UAS respectively. Possible decametric radio bursts have been
detected by BAS. Impact predictions: J. X. Zhang and his colleagues provided the July 7
edition of their heliocentric orbital elements and their impact prediction of SL9.
This work was mainly supported by Chinese Academy of Sciences and Chinese National
Science Foundation.
115

THERMAL EFFECTS FROM MINOR CONSTITUENTS
IN JUPITER'S UPPER ATMOSPHERE AFTER SL9
Y. Wang & K. S. Noll
Space Telescope Science Institute
3700 San Martin Dr.
Baltimore, MD 21218
R. V. Yelle
NASA/Ames Research Center
MS245­3
Moffett Field, CA 94035­1000
The thermal structure of Jupiter's atmosphere at pressures between 10 \Gamma6 mbar and 100 mbar
is calculated for both pre­impact and post­impact conditions, using a radiative­conductive
model. For the pre­impact model, we include solar UV and EUV heating, non­LTE cooling
by hydrocarbon fundamental bands, cooling by H 2 collision­induced opacities, and heating
by CH 4 near­IR and visible bands. The calculated temperature pro®le indicates that extended
heat sources other than known gas opacities have to be present in order to explain the observed
temperatures in both the stratosphere and the mesosphere of Jupiter. We added a total heat
#ux of 175 ergs/cm 2 /sec in the stratosphere, 15 ergs/cm 2 /sec in the mesosphere, assuming
a Chapman pro®le for each source. This heating is presumably related to the absorption
of solar energy by suspended aerosals. Some reports ®nd post­impact temperatures in the
stratosphere decreased by as much as several tens of degrees. To study the effects of impact­
generated gas on Jupiter's thermal pro®le we consider a range of models with stratosphere
abundances of HCN, NH 3 , and CO taken from published reports. Our studies show that
the cooling effects from the minor constituents including HCN, CO, and NH 3 , produced
by the impact, are insigni®cant. The existing thermal balance is dominanted by the main
hydrocarbon fundamental bands and H 2 collision­induced opacity, and can not be perturbed
signi®cantly by the observed quantities of impact­generated gases.
116

Geometrical Model for the Plume Evolution
J. Watanabe, S. Takeuchi, H. Hasegawa, B. Suzuki and K. Sekiguchi
A geometrical model of the evolution of the impact plume is described. Although impact
plumes were basically observed with the HST, three­dimensional structure of the plumes is
not clear. Considering the geometrical distribution of the dust grains on the stratosphere
at the impact sites, we constract the three­dimensional model of the plumes. By using this
model, we discuss the nature of the light curve of the impacts recorded by many grand­based
observations with Near­IR wavelength.
117

Temporal and Spatial Variations Among the Fragments of
Comet D/Shoemaker­Levy 9
H. A. Weaver (STScI), M. F. A'Hearn (UMD), C. Arpigny (Li# ege),
P. D. Feldman (JHU), Ph. Lamy (LAS), K. J. Meech (Hawaii),
K. S. Noll (STScI), & T. E. Smith (STScI)
Observations of comet D/Shoemaker­Levy 9 (SL9) with the Hubble Space Telescope began
in July 1993 and continued systematically throughout 1994. While a typical spatial brightness
pro®le can be de®ned for the SL9 fragments, there were also signi®cant deviations from the
average behavior. For at least some cases, the deviations are likely due to fragmentation
events subsequent to the breakup of the SL9 parent body. There are apparently secular trends
in the core brightnesses of some fragments, but generally the variation is less than 50%. All
of the observed fragments showed a strong stretching of the coma near the time of impact as
the differential acceleration across the coma due to Jupiter became signi®cant.
118

Comet Shoemaker­Levy 9: Support for the Rubble Pile Model
Paul R. Weissman
Jet Propulsion Laboratory
Pasadena, CA 91109
The observed behavior of comet Shoemaker­Levy 9 as it orbited and impacted Jupiter was
consistent with predictions of the primordial rubble pile (Weissman, 1986) and fractal (Donn
et al., 1985) models for cometary nuclei. The comet displayed repeated disruption events
along its orbit as sub­nuclei continued to separate from the individual fragments, and complete
fading and dispersal of some smaller fragments. Using the rubble pile model, Asphaug and
Benz (1994) demonstrated the reassembly of the disrupted comet into ¸15­20 fragments,
with physically reasonable densities of 0.5 to 1.3 g cm \Gamma3 . Impacting fragments deposited a
substantial fraction of their energy in Jupiter's upper atmosphere, consistent with a partially
dispersed rubble pile.
119

Narrow­Band Imaging of the Impact Sites on Jupiter
Dennis D. Wellnitz and Michael F. A'Hearn
University of Maryland
College Park, Maryland
Ralph Martin and Arie Verveer
Perth Observatory
Western Australia
During the month of July 1994, the University of Maryland/Perth Observatory team recorded
about 6000 images and spectra of Jupiter, observing from the Perth Observatory in Western
Australia. Jupiter was observable from about 1030 through 1630 UT each evening, and
we enjoyed an unusually good run of clear weather and near­arc­second seeing during
much of impact week. Some of the images were taken through narrow­band ®lters in
methane absorption bands, while others were taken in continuum or wide red ®lters; the total
wavelength range sampled was from about 500 to 900 nm. This data set permits us to look at
the re#ectivity and absorptivity of the impact spots as a function of height in the atmosphere
of Jupiter, and to follow the evolution of individual spots during July 1994. We may also
integrate the absorption of a spot over its extent and thereby obtain an estimate of the total
mass of the particles comprising the spot. The current status and results of this analysis effort
will be presented at the meeting as a poster paper.
120

Scattering Properties of Dust Associated with SL9­Jupiter Impact
N.C.Wickramasinghe
University of Wales, Cardiff, U.K.
K.S.Krishna Swamy
Tata Institute of Fundamental Research
Bombay, India
The collision of SL9 with Jupiter injected substantial quantities of cometary dust into the
Jovian atmosphere at several impact sites, inparticular at the G­site. Therefore the observa­
tion of radiation #uxes from the impact site as a function of the time could yield important
information on the nature of the dust. In our preliminary investigation, we have calculated
the changes in planetary albedos based on optically thin radiation transfer model of an atmo­
spheric haze with optical depth at 5500A and with the size distribution function n(a)=const.
a(­3.6). The laboratory measured n and k values for the organic material has been used. The
results indicate signi®cant drop in albedo in the UV, but the situation is dependent on the
dust loading, the grain parameters and on the composition. We hope to explore wide range
of models including multiple scattering.
121

Chemistry of the Shoemaker­Levy 9 Jovian Impact Blemishes:
Indigenous Cometary vs. Shock­Synthesized Organic Matter
Peter D. Wilson and Carl Sagan
Laboratory for Planetary Studies
Cornell University
Ithaca, NY 14853
We ®nd that the optical constants of the aerosols in the dark impact blemishes, as measured
by West et al. (1995), are a near­identical match to Murchison organic residue. This organic
matter could be either (1) indigenous to the impacting comet, or (2) shock­synthesized from
the Jovian atmosphere. A 1­km comet will produce an earth­sized blemish with Ü = 1 if 5­
15% of its volume is deposited into the atmosphere as particles with radii r=0.1­1.0 microns.
Based on our most analogous charged­particle irradiation experiments, the energy deposited
into the Jovian atmosphere by the same­sized impact is able to produce only enough organic
matter to reach Ü = 0:25. This estimate, however, could easily be in error by an order of
magnitude, so we cannot conclusively exclude the shock­synthesis mechanism. Given the
high degree of similarity between the aerosols and the organic residue extracted from the
Murchison meteorite, though, we conclude that the organic matter in the blemishes is most
likely debris from the impactors.
122

Observations of the Jovian Satellites During Comet P/Shoemaker­Levy 9's
Impact with Jupiter
L. M. Woodney, R. Meier, M.F. A'Hearn, and D. Wellnitz
University of Maryland
T. Smith, A. Verveer, and R. Martin
Perth Observatory
Once it was known that the impacts of the fragments of comet P/Shoemaker­Levy 9 would
occur behind the limb of Jupiter it was expected that re#ections off of satellites would be one
of the best ways to view the entry #ashes from Earth. However, the results of the attempts to
observe this effect were disappointing. To date, no one has reported a signi®cant detection
of a #ash, but there are several reports of possible #ashes. Each of these detections require
con®rmation from other observers to establish whether or not they are noise.
Our own data contain two of these possible #ashes. We observed Io, Europa and/or the
limb of Jupiter during the expected impact times of fragments D, E, and K from two sites in
Western Australia. At the Perth Observatory 61 cm telescope we used a CCD camera and
broad band ®lters, while at Mt. Singleton we used an un®ltered, high­speed, blue sensitive
photometer and a 14 inch Celestron. Our best photometer observations are from the time of
the D impact. These data also include our best possibility for an observed #ash. There is a
7% increase in the brightness of Io which lasts several seconds. There is a second possible
#ash at the time of impact of E, which is seen in both the photometer data and a CCD image,
but both detections could easily be explained by noise. To establish whether or not these
events are real, high signal to noise data from other observers is needed.
The luminosities that we derive from both possible #ashes are two orders of magnitude
greater than seen by Galileo for other events. This implies that the spectra of these #ashes
were either very different from those of black­bodies, or possibly that they were like those
of very hot black­bodies.
Our upper limit for the luminosity of impact K of 6 \Theta10 23 erg sec \Gamma1 ¯m \Gamma1 at 945 nm is only
a few times brighter than the peak observed by Galileo. Perhaps better data reduction will
reveal a #ash.
123

Physical Properties of Swarmed SL9 Fragments at Impact
Xingfa Xie and Michael J. Mumma
NASA/GSFC
Kevin M. Olson
George Mason University
We present results from an N­body simulation of the swarms made of tens­meter­sized
objects, beginning at apojove and ending seconds before impact. Our simulation shows the
swarm to be stretched into a long train along the direction of orbital motion, with a high mass
concentration at the center. The initial sizes of the swarms at apojove are estimated to be in
the range of 20­50 km for swarm masses of 10 13 \Gamma 10 15 g, based on the light curves obtained
by Galileo. Incorporation of particle collisions into our model is currently in progress.
124

Explanation of Flashes and Bright Flare in IR­Lightcurves of Shoemaker­Levy 9
Impacts: Comparison between Numerical Simulations and Observations
Takashi Yabe, Masakazu Tajima, Sho Sasaki, Yutaka Abe, Feng Xiao,
Takayuki Aoki, and Jun­ichi Watanabe
The strong IR emission and the preceding #ashes at Comet Shoemaker­Levy 9 impacts
into Jupiter are studied based on the numerical simulation of the explosion process. 2­D
and 3­D simulations of impact and explosion processes are performed using highly accurate
hydrocode ExCIPHER. Energy deposition at the shallower depth along an elongated trajectory
of cometary entry can generate the ®rst rapid rising ®reball as well as a subsequent broad
plume. Calculated time intervals are compatible with the secondary #ash and the strong
bright #are in IR­lightcurves. Obtained temperature evolution can explain the observed rapid
decay of the strong #are.
125

CCD Methane Band Modelling of A, H, L and R Impacts
Padma A. Yanamandra­Fisher, Jose Luis Ortiz and Glenn S. Orton
Jet Propulsion Laboratory, Pasadena, CA 91109
Time series of impacts (H, L, R) at 0.89 ¯m and (A, H, L) 0.945 ¯m indicate that the main
event or the major brightening of the plume occurred and is in phase with the peak of the event
as observed at 2.3 ¯m. However, the ®rst and second pre­cursors are not always evident in the
CCD methane band data. We are presently working on modelling these events to establish
the spectral energy distribution of the fragments. Our data set comprises of data from Calar
Alto, La Palma and Table Mountain observatories.
126

Three strong Jovian decametric radio bursts from impacts
Xi­Zhen Zhang 1 , Jin­Lin Han, & Qi­Bin Li
Beijing Astronomical Observatory
Chinese Academy of Sciences
Beijing 100080, P.R. China
1 email: zxz@bao01.bao.ac.cn
Three strong decametric radio bursts (?20db), undoubtedly related to the impacts of frag­
ments G, K, and W of Comet Shoemaker­Levy 1993e (SL­9) on Jupiter, have been detected
during the impact week at Xin­Xiang temporary Jovian Decametric Watch Station of Beijing
Astronomical Observatory. All of them are narrow band events occurred at about 26.0 and
28.5MHz, and have a long duration of at least several minutes. Their completely different
characteristics, (ie. strongest one before G impact, immediate response to K impact, and
switching bursts after W impact), are shown in this report.
The project is supported by the National Natural Science Foundation of China.
127

On Jovian Decametric Radio Bursts of the Great Comet Crash
Xi­Zhen Zhang , Lin­Lin Han, Qi­Bin Li , Pei­Nan Jiao, Jun­Hu Du
Strong decametric narrow­band radio bursts (N events) and broad­band storms, produced
by impacts of fragments of Comet Shoemaker­Levy 1993e (SL­9) on the Jupiter, have been
detected during our observations from July 17 to 22, 1994 at the Xin­Xiang temporary Jovian
Decametric Station (longitude = ­113.8 degrees, latitude = 35.3 degrees) of the Beijing
Astronomical Observatory. Observations were carried out in the scanning mode in the
frequency range from 24.0 OMHz to 28.5 MHz. Signals were detected from 10 narrow
bands in a step of 0.5 MHz, and each band has a bandwidth of 0.1MHz and a sampling time
of 0.1 second.
Plots of relative signal intensity of these bursts at 10 frequencies between 24.0 and 28.5 MHz
has been published (Zhang et al., 1994 and Han et al., 1994). In this paper we report some
preliminary analysis results on three strong radio bursts, which were apparently produced
by G, K and W fragment impacts on July 18, 19 and 22, respectively. Also some spectrum
analysis are presented on decameter storms which appears to be related to the impact of
fragment S=5 on July 21.
128

THE PREDICTION OF THE IMPACT OF COMET SHOEMAKER­LEVY 9 ON
JUPITER
J.­X. Zhang, Q. Wang, J.X. Yang, S.C. Wang, X.Z. Chen
Purple Mountain Observatory, Nanjing, China
Using numerically self­established dynamical model for solar system objects, exploring
method and making software, we successfully sent out several times of prediction from
April to July 1994, and the predictions mainly coincided with actuality. The method and
theory of the predictions are expounded brie#y in this paper, and the precicted impact
items are compared and analyzed with actual accented impact times newly con®rmed. The
nuclei's movement when they were in the vicinity of Jupiter and the impact positions are also
discussed.
129

THE LIGHT CURVES OF GALILEAN SATTELITES DURING THE COLLISIONS
OF COMET SHOEMAKER­LEVY 9 AND JUPITER
Hongnan Zhou
Dept. of Astronomy
Nanjing University
Nanjing 210093
P. R. China
In this paper, we present the light curves of Galilean Sattelites during the fragments D=19,
E=18, K=12, N=9, P2=8b and S=5 impact with Jupiter.
130

THE IMPACT TIME OF FRAGMENT P2=8b
Hongnan Zhou
Dept. of Astronomy
Nanjing University
Nanjing 210093
P. R. China
According to the light curve of Io during the fragment p2=8b impacts with Jupiter, we
obtained the precise impact time of fragment p2=8b.
131