Home > SESSION 01 ....... Comet Shoemaker-Levy 9 I

SESSION 01 ....... Comet Shoemaker-Levy 9 I

SESSION 01 ....... Comet Shoemaker-Levy 9 I
Monday, 8:30 - 10:00 Crystal Ballroom
M. Festou and H.A. Weaver, Moderators

01.01 Weaver H. A.* Noll K. S. Storrs A. D. Smith T. E.
A'Hearn M. F. Arpigny C. Feldman P. D. Boice D. C.
Stern S. A. Lamy P. L. Larson S. M. Levy D. H.
Scotti J. V. Marsden B. G. Meech K. J. Shoemaker C. S.
Shoemaker E. M. Sekanina Z.
HST Monitoring of Comet P/Shoemaker-Levy 9

The Hubble Space Telescope (HST) first observed comet P/Shoemaker-Levy 9
(SL9) on 1 July 1993, and then began a systematic monitoring campaign in
late January 1994 that continued through the impact week in mid-July
1994. During the campaign a total of 39 HST orbits were devoted to
imaging of the comet, and 6 orbits were devoted to UV spectroscopy with
the Faint Object Spectrograph (FOS) in the wavelength range from 2223 -
3278 nagstrom. The images were used to estimate the sizes of the nuclei,
examine carefully the near-nucleus morphology to search for further
fragmentation events and outgassing activity, monitor the temporal
variability, and measure the V-R colors in the inner comae of several
nuclei. Results from several different image deconvolution analyses
indicate that the diameters of the larger nuclei were in the range of
about 2 - 4 km (assuming a geometric albedo of 4%), but all of these
techniques impose constraints on the core behavior of the coma
brightness profile that are not necessarily valid. Most of the nuclei
displayed spherically symmetric comae of approximately 0."6 in radius
from July 1993 until mid-July 1994, at which time the comae became
highly elongated along the direction of the train. There is clear
evidence for continuing fragmentation of the nuclei subsequent to the
July 1992 breakup of the original body, which seems to argue against the
"swarm" model for most of the fragments. The spectroscopic observations
of SL9 have not shown any evidence for OH emission, which might be
expected if SL9 is a water-laden object. Spectra taken on 14 July 1994
show strong Mg II emission in one two-minute segment followed by a
three-fold increase in the continuum about 18 min after the appearance
(and disappearance) of Mg II. The continuum enhancement lasted for less
than 20 min. These latter spectral observations were made when SL9 was
about 50 R(sub)J from Jupiter and may correspond to the time when the
comet crossed the Jovian magnetopause.

01.02 Benner L. A. M.* McKinnon W. B.
Dynamical Evolution of P/Shoemaker-Levy 9

We have investigated the orbital evolution of P/Shoemaker-Levy 9 by
integrating its motion in the general four-body Sun-Jupiter-Saturn-comet
problem, using the orbital elements determined by D.K. Yoemans and P.W.
Chodas (posted on the P/Shoemaker-Levy 9 public account). We have now
simultaneously integrated nineteen fragments (posted February, 1994)
back to the 1992 perijove to test the effects of self-gravity (suggested
by Benz and Asphaug) and possibly constrain the masses of the fragments.
A general convergence within approximately 0.2 R(sub)J of the train
center shortly after perijove is observed, but in general, improved
orbital elements are necessary to achieve a more compact convergence, so
no constraint on the fragment masses can be inferred yet. Prior to
perijove, Shoemaker-Levy passed through the inner halo of Jupiter's ring
at approximately 1.6 R(sub)J. Collisions with ring particles may have
weakened the nucleus and assisted tidal disruption, or caused the
disruption as the comet is apparently not strength dominated. The motion
of each fragment was treated as a test case of the pre-disruption parent
body's evolution and was integrated backward. Capture durations vary
between approximately 6 and approximately 40 P(sub)J (P(sub)J = 1 jovian
year), with a median near approximately 9 P(sub)J. The motion of the
comet is chaotic, with a Lyapunov time of approximately 30 yr, so no
definitive capture date can be inferred. The only robust inference for
its pre-encounter heliocentric origin is that the orbit was not one of
either high inclination or high (>0.5) eccentricity. Secular solar
perturbations cause Shoemaker-Levy's jovicentric orbit to alternate
between intervals of low eccentricity (e similar to 0.1 - 0.3) and high
inclination (i similar to 90 degrees), and near-parabolic eccentricity
and lower (either prograde or retrograde) inclinations (while its
jovicentric semimajor axis and angular momentum perpendicular to
Jupiter's orbital plane are approximately constant). Consequently, prior
to disruption Shoemaker-Levy's perijove oscillated from a maximum of
approximately 0.15 AU to a minimum well within the orbit of Callisto,
indicating multiple earlier close flybys may have occurred. Ultimately
solar perturbations increased Shoemaker-Levy's eccentricity and
decreased its perijove sufficiently for disruption in 1992 and collision
in 1994.

This work supported by NASA grant NAGW-432 and NGT-50629.

01.03 Lisse C. M.* A'Hearn M. F. Esterle P. A. McFadden L. A.
Weaver H. A. Woodney L. M.
Modeling of UBVRIJK' Observations of P/Shoemaker-Levy 9:
Implications for the Dust Size Distribution, Emission History,
and Icy Composition

Comet P/Shoemaker-Levy 9 was observed in the UBVRIJK passbands from
January 1994 through its July 1994 impact with Jupiter from the Kitt
Peak National Observatory (KPNO), Cerro Tololo Interamerican Observatory
(CTIO), and the NASA/Infared Telescope Facility (IRTF). Extended comae
and dust tails many arc-seconds in extent were detected for 10 of the
cometary nuclei. Observations to date have yielded only non-detection
upper limits for gas production from the comet. But the observed pre-
impact dust in the tails, trails, and debris fan associated with 1993e
had to originate at some point in time and space, presumably at/near the
July 1992 perijove that first captured the comet into a Jupiter-centered
orbit and fractured the parent body. The lack of activity due to low
local equilibrium temperatures suggests that any observed dust has a
substantial icy component to its composition. The behavior of the dust
during the July 1994 impact will place strong constraints on its
properties, and demonstrate its behavior in a strong magnetic field

We have fit the UBVRIJK images to modified Finson-Probstein/Monte Carlo
dynamical models. These models calculate the classical orbit of a dust
particle experiencing the gravitational influence of the Sun and Jupiter
and the radiation pressure of the Sun. The observed dust tail is then
fit by adjusting the rate of dust emission vs time, the velocity of dust
emission vs. particle size (Beta), and the particle size (Beta)
distribution. The models are poorly constrained unless observations over
a large angular scale and a large time range (>1 month) are made. We
present preliminary conclusions concerning the emission history and dust
particle size distribution of P/Shoemaker-Levy 9 from the dynamical

We have also combined the images into spectra. We fit the spectra using
a Mie scattering code modified to allow for porous spheres, multi-
component dust, and various particle size distributions. Cross-checks to
the particle size distributions found by the dynamical modeling are
made. We present preliminary conclusions concerning the
silicate:carbon:ice composition of the grains.

01.04 Ahrens T. J.* Takata T. Friedson A. J. Orton G. S.
Radiative Signatures and Plumes from Impact of Comet
Shoemaker-Levy 9 on Jupiter

Pre-impact diameters of Comet Shoemaker-Levy 9 fragments (SL9) are
computed by comparing observed maximum height of plumes and their
radiance with detailed Smoothed Particle Hydrodynamic (SPH) modeling of
impact in three dimensions [Takata et al., Icarus, 109, 3-19, 1994;
Ahrens et al., GRL 21, 1087-1090 and 1551-1553, 1994a,b]. Comparison
between preliminary absolute peak radiances of the plume from impact of
fragment R observed by the NASA/Infrared Telescope Facility SL9 Team
with the MIRAC2 camera (Hoffmann et al., Infr. Phys. Tech. 35, 175,
1993) at 7.85, 10.3 and 12.2 microns with SPH power versus wavelength
yields (using logarithmic interpolation at 10 microns) a diameter of
approximately 1.3 km. In contrast, maximum plume height after impact is
directly proportional to fragment energy, for example, inferred from
images of the G impact recorded by the Hubble Space Telescope (Hammel et
al. 1994, Science, in prep.) yields a height above the ammonia cloud
deck of approximately 3300 km. Interpolating previously maximum plume
height calculated for 0.4 and 2 km diameter fragments yields a fragment
diameter of 1.7 km (assuming a density of 1 g/cm^3). Using these values
to calibrate the relative sizes of fragments [Weaver et al., Science,
263, 787, 1994] yields an estimate of the progenitor SL9 diameter of 3
to 3.4 km and a total impact energy of 2.5 to 3.6 x 10^30 ergs or 6 to 9
x 10^7 Mton of TNT equivalent impact energy for the entire SL9 fragment
chain. The SL9 progenitor diameter is less than the seven short-period
comet nucleii in Luu's [Pub. Astron. Soc. Pac. 106, 425, 1994] catalog.
Comparisons between other aspects of the SPH model, such as the
penetration depth, will be made as new observational data become

01.05 Schenk P.* McKinnon W. B.
Characteristics of Split Comets: The Fossil Record from Crater
Chains on Callisto and Ganymede

Crater chains on the surface of Jupiter's large moons Callisto and
Ganymede represent a unique sample of ancient, tidally split comets.
These impressively straight chains, or catena, can be up to 700 km long
and consist of between 6 and 30 impact craters. No other type of crater
clustering has been recognized. Characteristics of these chains closely
resemble those of P/Shoemaker-Levy 9 at the time of discovery. Positions
of craters within chains exhibit no more than one crater diameter
deviation from the best-fit line through the chain. The largest craters
are located in the center portion of the chain. The chains approximate
great circles but can have curves equivalent to 30 degrees colatitude
small circles, consistent with nonvertical impact of a linear chain on a
rotating sphere. A reexamination of the Voyager images and detailed
mapping indicates that there is a total of 8 such chains on Callisto
that are clearly unrelated to secondary cratering (3 additional chains
are classified as of uncertain origin). Assuming a mean surface age of 4
b.y. for Callisto, we obtain a frequency of 1 splitting event every 150
to 200 years. Three similar chains have also been identified on the
younger surface of Ganymede. We have begun a study of the detailed
morphology of these craters in order to assess the role of dispersed
comet fragments on cratering in the Jovian system, and to characterize
the internal properties of comets/asteroids and their fragmentation.
Craters within these chains range from 4 to 36 km in diameter. Using
appropriate scaling laws, and an impact velocity of approximately 15
km/s at Callisto, we calculate cometary fragment size range of 0.2 to
several km, assuming a density of 1 gm/cm^3. Crater morphology may also
provide limits on the structure of individual fragments. Central peaks,
rimwall slopes and flat floors are recognizable in all the crater chains
(within the resolution limits) and have dimensions consistent with those
of normal craters. This implies that split comet fragments are not
loosely bound disrupted clusters (like many secondary craters), although
they may range from solid cohesive fragments to closely bound rubble

01.06 Hammel H. B.* Dowling T. E. Harrington J. Mills J. R.
Beebe R. F. Simon A. A. Huber L. F. Ingersoll A. P. Toigo A.
Orton G. S. West R. A. De Jong E. Karkoschka E. Kuehn D. M.
HST Imaging of Jupiter's Atmosphere in July 1994: Dynamical
Effects from the Impact of Comet P/Shoemaker-Levy 9

In mid July 1994, fragments of the comet P/Shoemaker-Levy 9 plunged into
Jupiter over a period of several days. We used the Hubble Space
Telescope to image Jupiter before, during, and after the impacts to
search for atmospheric phenomena created during the collisions.

Many effects on Jupiter's atmosphere were detectable. For four separate
events (A, E, G, and W), fireball plumes were detected rising above the
limb of Jupiter; some of these plumes reached heights of 4000
kilometers. The plumes were visible in Jupiter's shadow, probably in
thermal emission; above the shadow, they reflected sunlight. We observed
several fresh impact sites, and detected transient expanding rings
around several sites, most likely caused by horizontal propagation of an
atmospheric wave. Estimates of the propagation speed will be reported.
Thin ejecta blankets sometimes extended up to 16,000 km from the impact
sites. The distribution of material was asymmetric; the patterns were
consistent with ejection back along the sloping path of the incoming
projectile. Several features evolved throughout the week of impacts due
to the jovian winds. We also imaged impact sites over a period of
several weeks to watch the long-term effects of atmospheric circulation
on impact-created debris.

In this talk, we will present an overview of the dataset and the results
from our program. Detailed interpretation of the data will be presented
in other talks in the session.

These NASA/ESA Hubble Space Telescope observations were supported in
part through grant GO-5624.08-93A from the Space Telescope Science
Institute (operated by AURA Inc. under NASA contract NAS5-26555).

01.07 de Pater I.* Graham J. R. Jernigan G. Brown M. Liu M.
Marcus P. Fomenkova M. Ingersoll A. Orton G.
Infrared Observations of the Comet SL9/Jupiter Impacts at Keck

We observed the SL9 impact with Jupiter on July 16-21 with the near-
infrared camera (NIRC) mounted on the W.M. Keck telescope. We focused
on imaging experiments, at wavelengths between 1 and 4 micrometers. Our
most significant observation is the continuous monitoring of the
fragment R event at 2.3 micrometers. Monitoring began 25 minutes before
the impact of R, and continued until 22 minutes after. The resultant
movie has approximately 8 frames per minute, and shows two flashes at
the approximate time of impact, followed by the plume rotating into
view. We will show and discuss the movie and lightcurve of the R event.
In addition, we will show images of Jupiter's disk at 7 different
infrared wavelengths which show the time evolution and vertical
structure in the atmosphere of the various impact sites. We further have
images of several cometary fragments (M, K, and W), and of Jupiter's
ring (during impacts C and M, and afterwards).

01.08-INV Belton M. J. S. Chapman C. R.* Johnson T. V. Heffernan C. M.
Klaasen K. P. Galileo Imaging Team
Galileo Imaging of the Shoemaker-Levy 9 Impact Sites

The Galileo Solid State Imaging SSI Camera was used to observe directly
the impacts of fragments K, N, V, and W. To capture any sufficiently
bright bolides, fireballs, or other L luminous phenomena with different
sensitivities and time resolutions, two new camera modes were
implemented -- an on-chip mosaicking of 2 1/3 sec. time-lapsed images
and a diagonal drift-scan mode. Brief samples of data from the K, N, and
W impacts have been returned to Earth. A luminous event has been
detected for K from approximately 10:24:15 to 10:25 UT, reaching a peak
brightness (in the near-IR) about 10% that of E total Jupiter (as seen
from the same vantage point, 50 deg. phase, 1.5 AU from Jupiter). An
event was observed for W, commencing at 8:06:12 UT, and lasting for at
least 2 more images taken at 2 1/3 sec. intervals, which reached a peak
brightness about 1% the brightness of Jupiter (green filter). Data
return for the N impact is not yet sufficient to determine whether or
not there is an indication of an event. We have decided not to return
data for V, which were taken in a less favorable observing mode. It is
not yet clear which parts of the impact phenomenology have been observed
for K and W (e.g. bolide or fireball) but we expect to have much
additional data available for at least the K impact by the time of the
DPS presentation. We will also describe plans for returning the
remaining data during the months extending to the end of January 1995.

01.09 Martin T. Z.* Tamppari L. Orton G. S. Claypool I. Travis L.
Galileo PPR Observations of Shoemaker-Levy 9

The Galileo spacecraft Photopolarimeter Radiometer (PPR), a hybrid
visual/thermal IR instrument designed primarily to measure properties of
the Jovian atmosphere, was employed for SL9 as a staring high-speed
photometer at 945 and 678 nm, taking advantage of Galileo's direct view
of the impact point. Jupiter subtended 0.6 mrad within the 2.5 mrad
circular field of view. The PPR was able to acquire data at times when
no other Galileo optical instruments could operate. The impacts of
fragments B, H, L, Ql, and S were observed for 41 m each with a sample
time of 0.23 sec. The H, L, and Ql events provided good light curve and
timing information. Initial flashes occurred at these times (for
eardlbased observers):
H: 1994/199 19:31:58 UTC
L: 1994/200 22:16:48
Ql: 1994/201 20:13:52
These data likely represent thermal emission from the brief initial pass
of comet fragments through the atmosphere, followed immediately by an
expanding and cooling "fireball." The Ql flash appeared 3x brighter at
678 nm than at 945 nm, with a faster decay at the shorter wavelength.
Initial temperatures above 10^4 K are implied.
(A graph of these events appears on hard copy.)

01.10-P Yeomans D. K. Chodas P. W.
Comet Shoemaker-Levy 9 in Orbit About Jupiter

The collision of comet Shoemaker-Levy 9 with Jupiter has focused
attention upon the cometary formation process and the history of
cometary motions within the solar system. At least some comets appear to
be extremely fragile structures. In addition to the Jupiter tidal
disruption of comet Shoemaker-Levy 9s on July 7, 1992, comet Brooks 2
was disrupted by Jupiters tidal forces in 1886. Although Brooks 2 was
not captured by Jupiter, temporary captures of periodic comets by
Jupiter are not particularly unusual. During the 1800-2200 time period,
the following periodic comets either have been, or will be, in temporary
orbit about Jupiter: Shoemaker-Levy 9, Helin-Roman-Crockett, Oterma,
Gehrels 3, Smirnova-Chernykh, Haneda-Campos, Helin (1987w), and Kojima.
Apart from the record close Jupiter approach of comet SL-9 on July 7,
1992 (1.3 Rj) and the collision in July 1994, the next closest Jupiter
approach was that of Gehrels 3 in August 1970. Although Gehrels 3 did
not pass closer than 3 Rj to Jupiter, its osculating perijove distance
for an epoch in February 1971 was temporarily well below Jupiters
radius. Although only comet Shoemaker-Levy 9 collided with Jupiter,
solar perturbations on comets in Jupiter orbit can easily reduce a
comets perijove distance to allow a collision. Using the orbital
history of Shoemaker-Levy 9 as an example, the Jupiter capture process
is discussed.

01.11-P Rettig T. W. Hahn J. Tegler S. C. Mumma M. J. DiSanti M.
Hubble Space Telescope Observations of Comet Shoemaker-Levy 9
Subnuclei: Solid Bodies or Swarms?

We present a preliminary analysis of comet Shoemaker-Levy 9 images taken
with the Hubble Space Telescope Wide Field (WF) and Planetary Cameras
(PC) during January, March, May, June and July 1994. The WF and PC
images have resolutions of 0.10 and 0.046 arc sec, respectively. The
high spatial resolution of the images allows us to probe the dust
distribution in the inner coma of each subnucleus through an examination
of continuum surface brightness distributions. We are using the
observed magnitude variations and surface brightness distributions of
the inner comae to determine whether the various impacting subnuclei
consisted of kilometer sized central bodies or swarms of meter-sized
cometesimals. We suggested (Rettig et al. BAAS 1994) that as the
original cometary nucleus approached perijove in 1992, the nucleus
fragmented into a cloud of cometesimals that was quickly stretched out
in a 'tube' along the tidal direction. After leaving the Roche limit,
the decreasing tidal force permitted self-gravity to re-focus material
to form (approximately 20) 'swarms'. Depending on swarm mass and
particle relative velocities, several types of comet 'fragments' may
have developed. The swarms might have gravitationally re-collapsed into
central subnuclei or they may have remained as loosely bound swarms of
particles that could have been disrupted at large distances from
Jupiter, prior to impact. For subnuclei that consisted of loosely
formed swarms of particles, the effects of tidal disruption were
predicted to be observable prior to Jovian impact as variations in both
surface brightness distributions and magnitudes. Even the recondensed
subnuclei were expected to disrupt again at the tidal limit and were not
expected to penetrate much below the stratospheric levels of the Jovian
atmosphere. Evidence of swarm-like structures will provide information
on the physical structure of comet nuclei and will provide constraints
on comet accretion processes that existed some 4.5 billion years ago.

This research is supported by an STScI grant to the University of Notre

01.12-P Rauer H. Osterloh M.
Low Resolution Spectra of Comet Shoemaker-Levy 9 (1993e)

The activity of cometary nuclei is found to be restricted to descrete
areas on the surface. The difference between active and inactive regions
may be caused by the evolution of the comet during its perihelion
passages, or may be linked to processes during its formation. Assuming
comets are made of several planetesimals, physical and/or chemical
differences between the cometesimals may cause the observed differences
in activity. Up to now, the question of homogeneity of cometary nuclei
could not be addressed in measurements of comets. The break-up of comet
Shoemaker-Levy 9 into a chain of individual nuclei offers the unique
possibility to study possible differences of parts of a cometary

Low resolution spectra (16 angstrom/pixel) of comet Shoemaker-Levy 9
have been obtained on April 15 1994 at the 2.2m telescope at ESO, La
Silla, using EFOSC 2. The spectral range covers 3600 angstrom - 9200
angstrom. The slit of 5.7 arsec length and 5 arcsec width was placed
along the train of nuclei. In this way 11 spectra of the train have been
obtained to search for possible differences in the dust colour of the
individual nuclei.

01.13-P Chen J. Jewitt D. Luu J.
Mauna Kea Observations of Split Comet Shoemaker-Levy 9

Split comet Shoemaker-Levy 9 has been observed from Mauna Kea at regular
intervals since its discovery in March 1993. Systematic observations
with a charge-coupled device on the University of Hawaii 2.2 meter
telescope show the photometric and morphological evolution of the
fragments over the 16 month pre-impact period. Apart from the
progressive separation of the fragments, the most prominent change
observed is the decrease in the amount of dust associated with the
comet. Individual dust tails became prominent in early 1994 but
subsequently faded. As we reported at last year's DPS meeting, no clear
evidence for embedded nuclei can be discerned in the Mauna Kea data.
Spectra from the 3.6 meter CFHT and 10 meter Keck telescopes were
obtained to search for gas (OH, CN, N(sub)2^+, CO^+) in the 3000
angstrom to 4500 angstrom wavelength region. No evidence for emission
was found. We will present an overview of the temporal development of
the fragments prior to their plunge into Jupiter.

01.14-P Peterson B. A. Wehinger P. A. Wyckoff S.
Spectroscopy of Comet Shoemaker-Levy 9 and Jupiter

Spectroscopic observations of Comet Shoemaker-Levy 9 and Jupiter were
acquired during the period 1994 July 16-23 using the Mt. Stromlo Obs.
1.9-meter telescope and a long slit CCD spectrograph. We shall compare
the spectra of the fragments of comet Shoemaker-Levy 9 with those of
P/Halley observed at similar heliocentric distances (4.8 AU).

In addition, we report on slit spectra taken at the eastern limb of
Jupiter in an attempt to search for evidence of the effects of the
collision of selected larger fragments in the outermost layers of the
Jovian atmosphere.

This research is supported in part by the NASA Solar System Exploration
Division and the NSF U.S.-Australian Cooperative Science Program.

01.15-P Chernova G. P. Birkle K. Kiselev N. N. Jockers K.
Imaging Photometry, Polarimetry and Color of Comet
Shoemaker-Levy 9

Comet Shoemaker-Levy 9 was observed with the ESO 1m-telescope and the
focal reducer of the Max-Planck-Institute for Aeronomy from April 27 to
May 1. During this interval the phase angle wa.s always less than 0.4
degrees. ,About 60 useful images of 5 minutes duration were obtained in
the g, r, and i filters of the Gunn system. The field of view was 14.5 x
9.6 arcmin and the scale 1.6 arcsec pixel^-l. For the night of April
27/28 the following Gunn g magnitudes have been derived (the fluxes are
taken in square boxes of 5 x 5 pixels, i.e. 8 x 8 arcsec, corresponding
to 21000 x 21000 km at the comet): 20.6 A(21), 20.0 B(20), 20.2 C(l9),
19.3 E(17), 19.6 F(16),18.4 G(15),18.8 H(14),18.5 K(12),18.9 L(ll),l9.1
P2(8b), 19.3 R(6), 18.9 S(5), 19.5 W(1). The Gunn G magnitude of P1, Q1
and Q2 together, measured in a square box of 11 x 11 arcsec is 17.7. The
colors of the individual comet fragments are the same within the error
limits. This indicates a homogeneous nucleus of the parent comet.

In the night from May 12 to May 13 comet Shoemaker-Levy 9 was observed
with the prime focus camera of the 3.5m-telescope of the German-Spanish
Astronomica1 Center, Calar Alto, Spain, through a rotatable polaroid and
a filter similar to the R filter of the Cousins system in order to
obtain a polarimetry map of the comet. At that time the phase angle was
2.6 degrees. Four images of 20 minutes duration were secured, with
polarization directions 0 degrees, 45 degrees, 90 degrees, and 135
degrees. The field of view was 7 x 7 arcmin and the scale 0.41 arcsec
pixel^-l. Since the night was not photometric it will not be possible to
extract an accurate polarization map but we might be able to judge about
polariza.tion differences between the individual nuclei and their tails
and to compare them with the color variations.

01.16-P Yabe T. Xiao F. Zhang D.-L. Sasaki S. Kobayashi N. Abe Y.
Terasawa T.
Effect of EOS on Break-up of Shoemaker-Levy 9 Entering the
Jovian Atmosphere

Break-up of Comet Shoemaker-Levy 9 entering Jovian atmosphere was
studied with highly accurate hydrocode ExCIPHER[1] by changing the
equation of state (EOS) of the comet[2]. The comet of radius 1.5km was
completely destroyed around the atmospheric pressure of 10 bar which is
shallower than that of the other simulations[3,4]. Furthermore, the
behavior of break-up is quite different from them. This behavior depends
on the type of EOS . For hard material like water and ice, the impact of
Jovian atmosphere caused incompressible surface wave propagating along
the comet surface thus enhancing ablation loss even at a low pressure
atmosphere. For ideal EOS, however, the surface wave did not appear
since the comet was compressed upon impact and hence the ablation loss
around 5 bar was quite low. Although the comet was broken into moderate
pieces at 10 bar, most of the energy had already been lost well before
the break-up. Surprisingly, however, the final disintegration occurred
around 10 bar for tested EOS and even for planar geometry. This may be
because the disintegration is triggered by a hole created along the
central axis and the comet is torn off when the atmospheric gas
penetrates through the comet causing turbulent now behind it. A hole of
this kind can easily develop because even a small concave structure can
grow into a larger one by the focusing of flow through conical geometry.

The subsequent explosion process was also simulated by the same code. In
the first 100 second, a plume followed by a fireball propagates above
1000 km and then the plume alone is ejected higher in the sky like an
umbrella. The rest part expands quite slowly over 10000 km and this
structure lasts for more than 1000 second. These two sequences might be
related to the two subsequent increases of observed infrared signals.

[1] Yabe, T. et al. (1994) Nucl. Des. Engnrg. in press. [2] Yabe, T. et
al. (1994) J.Geomag.Geoelec. 46, 657-662. [3] Mac Low, M. and Zahnle, K
(1994) Ap. J. Lett. in press. [4] Ahrens, TJ. et al. (1994) Geophys.
Res. Lett. 21, 1087-1090.

01.17-P Wyckoff S. Ferro A.
Spectroscopic Observations of Comet Shoemaker-Levy 9 and

Long slit spectra of selected fragments of Comet Shoemaker-Levy 9 and of
Jupiter were acquired with the 4.5-meter Multiple-Mirror Telescope using
the red channel spectrograph. Observations were obtained during the week
of the SL9's collision with Jupiter. The spectra of the fragments will
be characterized. We shall compare the SL9 spectra with those of
P/Halley observed at similar heliocentric distances (4.8 AU).

This research is supported in part by the NASA Solar System Exploration

01.18-P Festou M. C. Lelouch E.
On the Detection of Gaseous Species in Comet Shoemaker-Levy 9
Prior to Its Impact with Jupiter

Comet P/Shoemaker-Levy 9 was observed on five consecutive nights with
the 30m IRAM antenna, from 16 July till 21 July 1994. The instrumental
setting allowed to simultaneously monitor the gaseous emissions of CO
and HCN in the 2 mm window until the comet disappeared behind the planet
disc, then to switch to another observing program. If in July 1994 the
comet sub nuclei break as they did in 1992 after entering the Roche
limit of the planet, then a large cloud of dust particles will form and
slowly expand into space. Large amounts of gas might be released, either
at the time of breakup or immediately after, when large areas of fesh
comet snow are exposed to sunlight. Such an event could occur anywhere
during the last hour prior to the impact. At the time of impact, this
cloud might be up to about 1,000 km in diameter, and very bright,
especially in CO light. The observing geometry is such that the comet
activity at mm wavelengths can be monitored until the last few minutes
before the impacts, maximizing thus the chances to detect a gaseous
emission from the comet. This paper will be given only if positive
results are obtained.

01.19-P McFadden L. A. A'Hearn M. F. Esterle P. A. Lisse C. M.
Weaver H. A. Woodney L. M.
Multi-Wavelength Imaging of Comet P/Shoemaker-Levy 9:
January-July, 1994

We have acquired multi-wavelength images of Comet P/Shoemaker-Levy 9
from January-July using three telesopes for seven observing runs in
order to monitor the brightness of the fragments, the dust tails,
trails, and debris fan.

Wide-field (FOV = 1 degree^2) CCD images with UBVRI filters were
acquired with the 0.6m Schmidt telescopes at Kitt Peak National
Observatory (KPNO) February, April, and June, 1994 and Cerro Tololo
Interamerican Observatory (CTIO) in March and June, 1994. Moderate-field
(FOV=1 arcmin^2) near-infrared images with J and K filters were
acquired with the 3m NASA-Infrared Telescope Facility at Mauna Kea in
January, May, and July, 1994. The 256 x 256 InSb array camera, NSFCam
was used at the IRTF. The brightness of the nuclei are in the range 20-
21 magnitudes/arcsec^2. The data will be analyzed for temporal
brightness variations of the nuclei, tails, dust trails, and debris fan.
Comparison with the colors of the Sun, comets McNaught-Russell (1993v),
Takamizawa-Levy (1994f) and asteroid 23 Thalia, also observed with this
equipment, will be made. We anticipate finding evidence of ice grains
that can be seen using broad-band visible and near-IR imaging although
if icy grains and dust are intimately mixed, these observations may not
be an appropriate diagnostic of water ice.

This work was supported by NSF and NASA.

01.20-P Chodas P. W. Yeomans D. K.
Comet Shoemaker-Levy 9 Impact Times and Impact Geometries

The fragments of Comet Shoemaker-Levy 9 impacted Jupiter July 16-22,
1994, within 10 minutes of the final predicted impact times. Post-impact
estimates of the impact times appear in the table below. These estimates
are based upon direct observations by the Galileo spacecraft (for
fragments H, K, L, and W), the times at which the Hubble Space Telescope
observed impact plumes, the times at which infrared flashes were
observed from the ground, the longitudes of impact features seen in HST
images, and the relative impact times predicted from astrometry. The
table also gives l-sigma uncertainties in minutes.

Fragment Impact Time Fragment Impact Time
(UT) (UT)
A July 16 20:11 +/-3 N July 20 10:31 +/-4
B July 17 02:50 +/-6 P2 July 20 15:23 +/-7
C July l7 07:12 +/-4 Q2 July 20 19:44 +/-6
D July 17 11:54 +/-3 Ql July 20 20:13 +/-3
E July 17 15:11 +/-3 R July 21 05:33 +/-3
F July 18 00:33 +/-5 S July 21 15:15 +/-5
G July 18 07:32 _+/-2 T July 21 18:10 +/-7
H July 18 19:32 +/-1 U July 21 21:55 +/-7
K July 19 10:24 +/-1 V July 22 04:22 +/-5
L July 19 22:16 +/-1 W July 22 08:06 +/-1

The impacts occurred on the night side of the planet near the morning
terminator, varying from 9 degrees past the limb as seen from the Earth
for fragment A to 4 degrees for W. Fragment W passed behind Jupiter's
limb only 5 seconds before impact, at an altitude of only 200 km above
the l-bar level. The Jovicentric latitudes of the impacts ranged from
43.1 degrees S for fragment A to 44.2 degrees S for W. The longitudes of
the impacts relative to the midnight meridian varied more widely, from
64 degrees to 71 degrees. The velocity of the fragments at impact was
about 61 km/s, the incidence angle was about 43 degrees, and the azimuth
angle of the velocity vector was about 15 degrees west of north. The
time required for Jupiter's rotation to bring the impact sites into
Earth view varied from 19 minutes for A to 9 minutes for W.

01.21-P Asphaug E. Benz W.
Size and Density of Comet Shoemaker-Levy 9 Deduced by Modelling
Break-up of the Parent 'Rubble Pile'

The fragments of comet Shoemaker-Levy 9 produced a display more
spectacular than many dared to hope; equally surprising was the wide
variation in impact magnitude. In our prediction paper [1], a
heterogeneous chain of "rubble swarms" is created when a parent comet
with bulk density about 0.7 g/cc and grain density about 0.9 g/cc
suffers tidal disruption and self-gravitational reaccretion along the
1992 perijove trajectory. Higher density parents reaccrete into one
dominant clump, and parents with bulk density about 0.5 g/cc form about
20 clumps of nearly equal mass. We can explain the uniform crater
chains on Ganymede and Callisto and the recent impacts into Jupiter by
the same mechanism if the bulk density of comets is 0.6 +/- 0.1 g/cc,
with SL-9 slightly denser than average.

Many rejected this idea as soon as the first bolide struck, claiming
that strength was needed to penetrate the deep atmosphere -- in spite of
the fact that models predicting deep penetration [2] did not generally
include strength. Spectroscopic evidence eventually showed that the
impacts were, after all, quite shallow, bursting near the 1 bar level.
It was also claimed, in spite of calculations to the contrary [3], that
Jupiter's tide would disperse a strengthless bolide prior to atmospheric
entry. This is also not true. The final issue concerns the size of the
parent comet. Huge impact scars might seem to require very large
impactors, yet three models including our own [1,4,5] conclude that the
parent comet diameter was less than 2 km. We must strive to understand
how a comparatively small mass creates such a big blemish.

A typical rubble-swarm in our model is over 10 km in diameter before the
final encounter with Jupiter, with a loosely-packed (approximately 0.1
g/cc) "core" about 1 km across. These approximately 10^14 g swarms each
deposit approximately 2x10^27 ergs. While the core can be treated as a
single entity in airburst models, the outer swarm must be treated as
many independent impactors. A 1 km 0.1 g/cc bolide explodes near the one
bar level [6], whereas a 100 m approximately 1 g/cc grain explodes about
100 km higher, two seconds earlier. Hundreds of smaller bursts, each
depositing 10 to 100 Mt over a wide area before the approximately 3 Gt
explosion of the central bolide, could play an important role in
fireball and plume evolution.

[1] Asphaug, E. and Benz, W., Nature 370 1994. [2] Takata, T. et al.,
Icarus, in press. [3] Yanagisawa, M. and Konno, T., LPSC XXV 1994.
[4] Solum, J.C., Nature 370 1994. [5] Scotti, J.V. and Melosh, H.J.,
Nature 365 1993. [6] Zahnle, K. and Mac Low, M. M., Icarus 1994.

01.22-P Barker E. S. Cochran W. D. Cochran A. L. Na C. Pryor W.
Baines K. H. Briley M. Smith V.
Spectrophotometry and High-Dispersion Spectroscopy of Jupiter's
Southern Latititudes During the Impact Period

We report on visible spectrophotometry and high dispersion spectroscopy
taken of Jupiter during the impact period (17 July - 23 July 1994 UT).
Spectrophotometric observations were obtained using the longslit CCD
spectrograph (ES2) on the 2.1m telescope to cover the 0.55 to 1.08
micron wavelength range. A seeing limited spatial resolution of about 1-
2 arcsec was achieved resulting in a spectral resolution of about 10
angstrom. The spectrograph slit was rotated to allow spatial sampling
along and adjacent to the impact latitude (-44 degrees S). This scenario
allowed sampling of impact regions and unperturbed regions as functions
of both time and location. Spectrophotometric calibration of the low
resolution dataset via standard stars and solar analogs will be used to
provide absolute calibration of the narrow-band CCD images taken
concurrently in the 6190 angstrom and 8930 angstrom CH4 bands by the
McDonald Team on the 0.76m telescope (see adjacent poster paper) .

The high resolution spectroscopy was done on the 2.7m telescope with a
cross-dispersed echelle spectrograph, giving nearly complete spectral
coverage between 0.40 and 1.05 microns. The coude slit was 1.2 x 8.1
arcsec permitting spatial resolution of the impact latitude in either a
N-S or E-W direction. An R=60,000 allowed resolution of individual
rotational CH4 lines in the impact region spectra.

We will report on changes in the morphology of the continuum and the CH4
bands between 0.6 and 1.0 microns. For example, the observed impact
regions appeared as bright features only in the strongest CH4 absorption
regions within the 8930 angstrom band and conversely as dark features in
the weaker absorption regions. Both the low- and high- dispersion
spectra covered a range of CH4 band or line strengths which will aid in
the determination of the altitude distribution of the condensates formed
as a result of the impacts.

01.23-P Cochran A. L. Baines K. H.
Barker Barker E. S. Cochran W. D.
Frueh M. Kim S. Kim Y. Kudari A. Lester D. F. Na C.
Pryor W. Gyorgyey Ries J. Trafton L. Whipple A. L. Wiant J.
A Synopsis of McDonald Observatory Imaging Observations of
Jupiter Before, During, and After the Shoemaker-Levy 9 Impact

We report on optical CCD and IR images of Jupiter which were obtained at
McDonald prior to the impacts of Shoemaker-Levy 9, during impacts, about
one week following the last impact, and almost one month after the last
impact. The CCD observations were obtained through narrow-band filters
which isolate the 6190 and 8900 Angstrom CH4 bands and the adjacent
continuum regions at 6040, 7640 and 8290 Angstrom. A barlow lens was
utilized to expand the image on the chip so that Jupiter mostly filled
the chip. The CCD observations before the impacts were obtained on the
0.9-m telescope, while the CCD observations during and after the
impacts were obtained with the 0.8-m telescope. During the impacts and
the following time, we obtained at least some data on each of 15
"nights" (we started observing as early as 4pm). The IR images were
obtained on the 2.7-m telescope using a NICMOS3 array (Rokcam) with
filters to isolate the 1.5 micrometer NH3 band, the 2.3 micrometer CH4
band, the 2.12 micrometer H2 S(0) pressure induced dipole absorption and
the continuum at 1.58 micrometer and short K-band. Rokcam observations
were obtained on 11 nights during and following the impacts. In
addition to the CCD and Rokcam observations, we obtained about 14 hours
of real-time video observations using the Lunar Laser Ranging 0.8-m
telescope. In the course of our observations, we observed the flash
associated with the impact of fragment R. Observations were obtained
during several other impact periods but no flashes were detected. We
compare images of the planet obtained with the CCD and Rokcam, including
a movie in some wavelengths. We also show a time sequence of certain
longitudes of Jupiter. Cross-cuts of the images for different
wavelengths are compared to show how the morphology of the spots, bands,
etc. change with latitude and wavelength.

01.24-P Lupishko D. F.
The Results of the Ukraine Observation Program of
Shoemaker-Levy 9 Collision with Jupiter

The photometric, polarimetric, spectral, photographic and radio
observations of Jupiter before, during and after the collision are
carried out at the observatories of Ukraine, Middle Asia and Northern
Caucasus. The spatial and time variations of polarization
characteristics of atmosphere aerosol, the photometric profiles and the
moments of flashes in Jupiter atmosphere, the changes in the structure
of atmosphere cloud layer, the data on the state of Jupiter
magnetosphere during the collision and Jupiter photographs of high
spatial resolution (0.3 - 0.5 arc sec)are supposed to be obtained. All
the results are going to be reported.

This work was supported by grant of National Space Agency of Ukraine.

01.25-P Kleshchonok V. V. Churyumov K. I.
Photoelectric Observations of Io and Europa in the Course of
Fall Down of Secondary Nuclei A, H, Q2, Q1 and T of Comet
D/SL-9 on to Jupiter

Observations of the Jupiter satellites within the program for
investigation of the comet D/SL-9 - Jupiter collision were held with the
electrophotometer installed on the AZT-14 (D=0.5m) telescope at the
Lesniki station of the Kiev University Astronomical Observatory. The
electrophotometer worked in the regime of photon counting with
registration on the PC IBM. In the course of observations standard UBVR
filters were used. Temporal control tie was hand-made according to exact
signals time. Circumstances for observations did not reveal changes in
the satellites brightnesses which could be linked with reflection of
light of fireball in the Jupiter atmosphere, with the exception of the
nucleus fall down Q2. Here, two events that can be tied with the
outburst during the nucleus fall down were found out. The first event
has the following parameters: filter B, the outset of the outburst at
19:32:09 UT, continuity of the outburst 2.7 sec, the amplitude 0.12 mag
[l]. Parameters of the second event: filter B, the outset at 19:48:10
UT, the continuity - 1.0 sec, the amplitude 0.11 mag. The second flash
seems more probable both in continuity and it being close to the
predicted time of the nucleus fall down Q2 on to Jupiter [2]. 1.
Churyumov K.I. Io brightens. The Siderial Times, 1994, No. 10, p.5, IAU
XXIInd General Assembly, The Hague. 2. Yeomans D.K., Chodas P.W. Comet
SL-9 impact times. Interoff. memorand., 1994, 314.10-87.

01.26-P Larson S. Scotti J. Howell E. Marcialis R. Nolan M.
Winsiewski W. A'Hearn M. Wellnitz D. Campins H.
Rabinowitz D. Butner H. Tapia S. Vilas F. Doherty P.
Bessell M. Brosch N. Liebowitz E. Gilmore A. Kervin P.
Africano J. Levine S. Schuster W. Martin P. Verveer A.
Moreno F. Molina A. Rutten R. Kidger M.
The Comet Impact Network Experiment (CINE)

The Comet Impact Network Experiment (CINE), an international
collaboration designed to observe phenomena associated with the impact
of Comet Shoemaker-Levy 9 with Jupiter was successfully deployed at the
4.2-m WHT at La Palma, Spain; the 2.5-m at Las Campanas, Chile; the 1.5-
m at San Pedro Martir, Mexico; the 1.2-m of the Air Force Maui Optical
Station, Hawaii; the 1.0-m at Mt. John Obs., New Zealand; the 1.0-m at
Siding Spring Obs., Australia; the 0.6-m telescope at Perth Obs.,
Australia; a 1.0-m Air Force GEODSS telescope on Diego Garcia; and the
1.0-m telescope at the Wise Obs., Israel. Coronagraphic/spectrographic
transfer optics were constructed to reduce the scattered light from
Jupiter and normalize plates scales with the various facility CCD
cameras. The main objectives were to 1) provide astrometric positions of
the nuclei just prior to impact, 2) look for evidence of cometary
particle fragmentation as the comet enters the Jovian magnetic field, 3)
look for evidence of further breakup as the comet crosses its Roche
limit, 4) obtain time-resolved spectra of the meteor flashes reflected
off the satellites, 5) look for evidence of material ejected from the
impact, 6) observe changes in the Jovian cloud morphology, and 7) look
for evidence of dust capture and interaction with the Jovian ring. The
project was implemented through the local observer-collaborators who
applied for telescope time and provided the technical information needed
to interface the coronagraphs with the telescopes and CCDs, and
observers who accompanied the coronagraphs to the sites.

This project was supported by the NASA Planetary Astronomy Program, the
USAF Space Command and the various host institutions.

01.27-P Wellnitz D. D. A'Hearn M. F. Martin R.
Comet SL9 Interactions with Jupiter as Observed from Perth,

The results of planned observations of the interactions of Comet
Shoemaker-Levy 9 with Jupiter, using the University of Maryland CCD
Imaging System for Comets (CISCo) on the Lowell Telescope at the Perth
Observatory in Western Australia, will be reported.

Planned observations include coronographic imaging of the cometary
fragments as they approach and impact Jupiter, fast CCD imaging of
satellites of Jupiter near the predicted times of cometary fragment
impacts, imaging of the atmospheric fireball if it rises above the limb
of Jupiter, coronographic imaging of the Jovian ring to observe the
interaction of the cometary dust tail with the ring, imaging of Jupiter
with various filters in the continuum and in methane absorption bands of
various depths to search for cloud formation in the stratosphere after
the fireball starts to cool, and low-resolution spectrographic imaging
of Jupiter to search for changes in the absorption features in
appropriate areas of Jupiter.

This work is being supported by NASA and NSF funds.

01.28-P Jockers K.
Observations of Jupiter's Atmosphere in the K-Band During the
Shoemaker-Levy 9 Encounter

Using ESO's infrared camera IRAC2B at the German/ESO 2.2m telescope at
La Silla, Chile, observations of Jupiter's atmosphere were performed
from July 16 to 24, 1994 (8 nights). The IRAC2B camera employs a NICMOS-
3 array with 256 x 256 pixel of a size of 40 microns. A scale of 0.27
arcsec pixel^-l was used. Regrettably not a single impact was observed.

Filter images were taken through two interference filters with center
wavelengths 2.105 and 2.365 microns, and FWHM 0.037 and .088 microns
respectively. These wavelengths are located at the flank and in the
centre of a deep methane absorption band. While at 2.105 microns some of
the Jovian clouds are still visible, Jupiter's disk is practically black
at 2.365 microns and even the Great Red Spot, which extends to great
height, is barely visible. In both filters the impact clouds appear in
emission and have about the same brightness. This indicates that the
clouds must be located at atmospheric heights unaffected by methane
absorption. They are remarkably stable but in the course of a few days
are getting sheared by Jupiter's velocity field.

Narrow-band images of spectral resolution approximately equal to 1000
were obtained with an imaging Fabry-Perot-Interferometer in the "hot"
band of H(sub)3^+ at 2.093 microns and in the H2 quadrupole line at
2.121 microns. Both lines are sensitive to temperature enhancement in
the outer layers of the Jovian atmosphere. Usually the "hot" band of
H(sub)3^+ can be observed only when the auroral heating is particularly
strong. Quick look data show no evidence for the presence of the

The paper will present the time evolution of the impact clouds as
observed with the interference filters. In addition we will provide
evidence for the detection or non-detection of tlle H(sub)3^+ and H2
quadrupole line in tlle interferograms.

01.29-P Klavetter J. J. Petmecky S. Levine S.
Near-IR Reflection Observations of the Shoemaker-Levy 9

We observed the SL9 impacts into Jupiter from the San Pedro Martir
Observatory (SPMO), administered by the Observatorio Astronmica Nacional
(OAN) located in Baja, Mexico. This presentation will concentrate on
the near IR observations made from the 2.1 meter telescope using the
Camila IR imager.

We will report on the dramatic impact spots/plumes observed from the
SPMO/OAN site, including the L and R (and perhaps M and W) fragments.
We will show the dramatic increase in brightness from our time series
imaging, including photometry of the spots/plumes.

The morphology of the spots is dramatically different at the wavelengths
observed, from 1.1 microns to 2.3 microns, with the spots being
essentially invisible at the smaller wavelength bands. We will show
images from filters throughout this range and place temperature limits
on the thermal continuum radiation observed.

01.30-P Wasserman L. H. Buie M. W. Millis R. L.
Imaging of Shoemaker-Levy 9 Impact Phenomena from Northeastern

We will be conducting imaging observations of the Shoemaker-Levy 9
impact on Jupiter from Northeastern Austrailia from 1994 July 12-25. Our
observations will be made with our portable Celestron 14" telescope with
the Occultation CCD camera system that was presented at last year's DPS
meeting. The system has now been retro-fitted with an automated 10
position filter wheel and a new optical mode for an image scale of
approximately 0.5"/pix.

Our two primary goals are (1) to take extensive sets if images to
monitor the Jovian atmosphere for any changes caused by the impacts, and
(2) to attempt to observe the flashes from the impacts in reflection off
of a well-placed satellite. Current predictions indicate that from our
location, four impact occur in darkness. One of these is the only impact
that occurs with Europa in eclipse where a flash may be seen.

Our system can quickly cycle through the filter set for multi-spectral
imaging of Jupiter. At the time of fragment impacts, we can quickly re-
configure and take high-frame rate data (up to 5 Hz) for the flashes. We
have also developed a spectroscopic mode that can be used to derive the
color temperature evolution of the prompt flash, provided enough signal
is generated from the impacts.

We will present results from our efforts and will provide a catalog of
the dataset for interested people. This work was funded by NSF grant

01.31-P Lazzaro D. Moreira J. K. Foryta D. W.
P/Shoemaker-Levy 9 Impact: Observations at the Pico-dos-Dias

We report the observations planned at the Observatorio do Pico-dos-Dias
(Itajuba, Brazil) for the impact on Jupiter of comet Shoemaker-Levy 9
(SL9) expected to occur on July 94 (Marsden, 1993, IAU Circular No.
5801; S Green, 1993, IAU Circular No. 5807).

Since the impact of the 20-22 fragments of SL9 are expected to take
place on the side of Jupiter away from the Earth, we plan to observe the
reflected light of the bolide and/or fireball, subsequent to each
impact, from a suitably placed Galilean satellite.

A rapid photometer mounted on a 60cm telescope will probably be able to
detect these phenomena using an integration time of 0.1 seg. With this
instrument we expect an impact detectability limit of 14.5 to 12.5
magnitudes (V), whether or not the observed satellite is in eclipse.
Assuming a luminosity efficiency of 3% we estimate that the above
magnitudes correspond to a fragment diameter of 100-200m. A lower
luminosity efficiency will necessarily increase these diameters up to 1-
2 km, which would be just the greatest fragments.

The final aim of these observations is the obtention of some data
relevant to determine the fragment distribution of SL9. We pretend, by
the time of the meeting, to carry out at least preliminary reductions in
order to deduce the size of the fragments from the observed signals (if

01.32-P Glenar D. A. Hillman J. J. Le Louarn M. Drummond J. D.
Narrow-band Visible and IR Imagery of Jupiter During the Comet
Shoemaker/Levy-9 Collision Series

We observed the Southern hemisphere of Jupiter on several nights during
the Jupiter SL/9 impact sequence, using the Coude focus at the Kirtland
Air Force Base, 1.5-meter adaptive optics telescope. The objectives were
to identify condensed volatiles in the high atmosphere at the locations
of the impact sites, track the motions of these sites as they transit
the disk, and examine their breakup. Multispectral images were acquired
in the CH4 bands at 725 and 890 nm, the NH3 bands at 645 and 790 nm and
in their adjacent continua using a tunable acousto-optic imaging
spectrometer and silicon CCD at the focal plane, with a spectral
resolution (Lambda/Delta Lambda) of 380 to 530. The stratosphere near 10
mbar was simultaneously sounded by imaging in the deep 2.35 micrometer
CH4 band using a NICMOS-3 camera with cooled narrow band filter. All
observations were photometrically calibrated and reduced to absolute
reflectivity (I/F) using close coincidence standard stars and published
solar intensities. A summary of the major results is presented.

This work was supported by the NASA Planetary Instrument Definition and
Development Program and the GSFC Directors Discretionary Fund.

01.33-P Woodney L. M. Meier R. A'Hearn M. F. Wellnitz D. Smith T.
High-speed Photometry of the Jovian Satellites During Comet
P/Shoemaker-Levy 9's Impact with Jupiter

At Mt. Singleton, West Australia, a high-speed photometer and a 14 inch
Celestron were used to look for flashes off Io or Europa during the
expected impact times of D, E, K, and N. The observations were made in
broad-band white light. Io was well monitored during the impact time of
D and no flashes are apparent. There is a possible effect at the time of
impact of E, but due to problems with tracking caused by high winds, we
can not yet rule out contamination from Jupiter in the aperture.
Comparison with results from other observers may resolve this
uncertainty. No flashes were observed for the impact of fragment K,
which occurred while Europa was in eclipse, or for fragment N.

The brightness of the monitored satellites changed by less than
approximately 10%. Careful analysis should allow us to look for small
effects and better constrain upper limits.

01.34-P Stuwe J. A. Schulz R. Encrenaz Th. A'Hearn M. F.
Shoemaker-Levy 9 and Its Impact sites on Jupiter

We report on observations in the visible spectral range of P/Shoemaker-
Levy 9 before and Jupiter after the impact. SL-9 was observed with the
ESO Multi Mode Instrument (EMMI) at the 3.5-m New-Technology-Telescope
(NTT) at La Silla, Chile, on July 1 and 2, 1994, two weeks before its
impact on Jupiter. Highly resolved images of the entire cometary train
were obtained in B, V, R, and I. A comparative study of the individual
fragments in terms of color, brightness and morphological appearance
will be presented. Due to the large field of view of the instrument all
fragments could be imaged in one frame. The comparison of the individual
fragments relative to each other is therefore not affected by errors due
to changing observational conditions. The large spatial scale
(approximately equal to 1200 km pixel^-1) on the other hand allows the
detailed analysis of the dust distribution around the individual sub-
nuclei. The dust distribution around some fragments appears to be
elongated slightly towards the direction of Jupiter. Additionally, we
obtained long-slit spectrophotometric observations of SL-9 on both dates
using EMMI in the dichroic medium dispersion mode (DIMD), which allowed
the parallel use of a blue- and a red-sensitive CCD covering a spectral
range of 3200 angstrom less than lambda less than 10400 angstrom (3.5
angstrom pixel^-l). The slit was aligned along the cometary train from
fragment K to fragment S. The spectra of fragments K, L, P, Q, and S
were searched for the CN (0-0) emission band at 3875 angstrom. No CN
emission feature was detected for either fragment and we therefore
present 3-sigma upper limits on the CN production rates.

After the impact Jupiter was imaged on two nights (July 23/24 and 25/26,
1994) with the NTT and the Superb Seeing Imager (SUSI) in six different
filters to study the appearance of the impact sites in different
spectral ranges. The time difference of two days (approximately 5
Jupiter rotations) between both observational series also allows the
investigation of the temporal evolution of the impact sites.

01.35-P Rogers J. H.
Jupiter's Visible Clouds Following the Comet Impacts

Imaging and visual observations of Jupiter after the comet impacts
showed extremely dark spots, which were the darkest features on the
planet and in some cases grew larger than the Great Red Spot. The size
of each visible 'scar' was roughly in proportion to the reported
magnitude of the fireball. The scars always appeared at least as dark
near the limb as when fully on the disk. For example, sites H and Q1
were observed coming onto the disk one hour after the impacts, and
became darker on subsequent days, with an apparently black core.

There were no systematic motions (less than ten degrees per month)
relative to the S.S.S. Temperate Current or System III. There was
evidence for diverse local motions, viz: increase then decrease in
longitude of site E; eastward and westward expansion of site K/W;
northward extension of site H.

Most sites have lasted over a month, although site A disappeared within
2 weeks. Most sites were beginning to fade after 10-20 days, while grey
patches were appearing outside the previous boundaries of some sites.
By late August, sites A and C had disappeared, and sites Q1, H, and E
were reduced to indistinct condensations on a dusky belt which extended
all round the planet at the impact latitude. The three largest complexes
-- K/W, L, and D/G/S -- were still very large and dark in late August
but had changed shape; dark material had spread for tens of degrees
preceding the original boundaries of K/W and L.

Before the impacts, five dark patches around 40 deg.S had been tracked
in the S.S. Temperate Current. They appear to be part of compensatory
changes south of the STB following its revival in 1993. These dark
patches remained visible as the impact sites faded.

This report is the result of observations by amateur observers in
several countries, especially: F. Balella (Italy), M. Bosselaers and
colleagues (Belgium), D. Bruton and R. Schmude (USA), R. Bullen (UK), E.
Colombo (Italy), P. Devadas and Mrs. Komala (India), M. Foulkes (UK), H.
Gross (Germany), C. Hernandez (USA), J. Lancashire (UK), R.J. McKim
(USA), H-J. Mettig (Germany), H. Miles (UK), I. Miyazaki (Japan), T.
Platt (UK), J. Rogers (UK), G. Teichert (France), D. Troiani (USA), J.
Warell (Sweden).

01.36-P Phillips C. B. Mueller B. E. A.
Broad-Band CCD Photometry of Jupiter's Impact Scars

Observations of Comet P/Shoemaker-Levy 9's impact on Jupiter were made
using the 36-inch telescope on Kitt Peak on July 20 and 22 (UT), during
the impacts, and on August 3 (UT) as a follow-up night. Images were
taken using a Harris UBVRI filter set and a 3.5 magnitude neutral
density filter, which was placed directly in front of the filter wheel.
The telescope was set to f/13.5 which resulted in a resolution of
0.384"/pixel, and the images clearly show the dark impact scars.

Preliminary analysis has resulted in color composites which show both
structure within the spots, and changes in color due to the spots' aging
between the July and August observation dates, as the August follow-up
date was chosen to include the same rotational phase of Jupiter. We are
also obtaining relative photometry of the impact scars in relation to
Jupiter's Galilean satellites and the undisturbed northern hemisphere of
the planet. We hope to trace the brightness evolution of the spots to
provide a quantitative counterpart to the qualitative visual spreading
of the impact scars during the two weeks spanned by our observations.

01.37-P Moreno F. Molina A. Ortiz J. L. Munoz O. Kidger M.
del Rosario J. C. Larson S. Campins H.
Consequences of the Comet Shoemaker-Levy 9 Impact on the Jovian
Atmosphere from UV to Near-Infrared Imagery

The impact sites of the fragments of the comet Shoemaker-Levy 9 in the
Jovian atmosphere were studied at UV through near-infrared imagery at
various telescopes in Spain. Prior to the impact sequence, systematic
observations were obtained in order to get a Jovian atmospheric
reference. Methane band imaging at the center of the absorption features
of different strengths will help to study the height dependence in the
aerosol content and dynamic properties of the stratosphere.
Particularly, we present here some preliminary results from CCD images
obtained using the 4.2 m William Herschel Telescope at La Palma. These
images were obtained during very good sky conditions and sub-arcsec
resolution on 16-22nd of July. Low and high resolution (0.34
Angstrom/pixel) spectra were also obtained with the WHT using the ISIS
spectrograph. These observations have been made as a part of the Comet
Impact Network Experiment.

01.38-P McGhee C. Gierasch P. Goodman J. Hayward T. Moersch J.
Nicholson P. Squyres S. Van Cleve J. Matthews K.
Neugebauer G. Weinberger A. Orton G.
Observations of the Shoemaker-Levy 9 Impacts from Palomar Using
the 256 x 256 InSb Camera

We observed the effects on the jovian atmosphere of the collision of
comet Shoemaker-Levy 9 in July and August 1994, using near-infrared
imaging at the Palomar 200-inch with the Caltech near-IR Cassegrain
camera. This camera permits imaging through broadband J, H, K, or L'
filters, or two narrowband (R = lambda/delta lambda about 70) circular
variable filters tunable between 1.5 and 5.2 microns. Our observations
focussed on the spatial distribution and temporal evolution of the
impact sites, and their near-infrared spectral characteristics.
Observations were made from 15-27 July UT and from August 16-19 and 28-
29. On each night after July 21, several complete sequences of broadband
and narrowband images at 17 selected wavelengths in the 2.0-2.4 micron
and 3.0-4.0 micron regions were obtained. In narrowband images of
Jupiter in the strong methane bands at 2.2-2.4 microns and 3.2-3.5
microns the reflected light from the main tropospheric cloud deck is
almost competely suppressed, revealing the signatures of the
stratospheric hazes. Observations at 2.0-2.35 microns of the increasing
strength of the CH4 absorptions provide an altitude estimate of about 10
mb for these hazes. The 2.3 micron images from July 24-25, August 17-19,
and August 28-29 have been assembled into mosaics which show the
complete distribution of the impact hazes as well as their temporal
evolution. The longitudes of nine distinct features, associated with
impacts G, R, Q1, H, E, A, C, K and L have been measured in the July
data, and found to be consistent (within 3 degrees) with impact site
longitudes reported by Ingersoll et al. from HST images. With the
exception of R, all features remain identifiable in the August data,
albeit considerably altered in shape and size. By late August, sites G,
H, and E extended over about 20 degrees in longitude, while Q1, C, K and
L retained compact cores. These well-defined features show a mean
eastward drift relative to System III of about 13 degrees over 24 days,
or 5.5 ms^-1. The more diffuse structures appear to map the zonal wind
profile in the middle stratosphere. This work was supported by a grant
from the NASA Planetary Astronomy program.

01.39-P Foust J. A. Hammel H. B. Elliot J. L. Meech K. J.
Patten B. M.
Observations of the Impact of Comet Shoemaker-Levy 9 Using a
High-Speed CCD Camera

Although the fragments of comet Shoemaker-Levy 9 impacted on the far
side of Jupiter as seen from Earth, two phenomena associated with the
impact events were postulated to be visible: the impact flash, reflected
off an appropriately placed satellite such as Io; and a plume or
fireball from the impact site that rises beyond the Jovian limb and into
direct view from Earth. Both of these phenomena require observations
with high time resolution and accurate timing to provide details on the
depth and magnitude of the explosions.

We observed the impacts and Jupiter using a high-speed CCD camera
mounted on the University of Hawaii Planetary Patrol 0.61-m telescope at
Mauna Kea Observatory. This system is capable of taking five or more
images per second over a small field and utilizes a Global Positioning
Satellite (GPS) receiver to provide accurate timing, making this an
ideal system to carry out observations of impact flashes and plumes.

Of the four impacts potentially visible from Mauna Kea, weather
conditions permitted us to observe two, impacts C and R. High-speed
imaging of Io, Europa, and Ganymede during the time of impact C showed
no visible change in the R band brightness of the satellites. High-speed
imaging of the limb of Jupiter in B band during impact R showed no
strong signal for a plume or fireball at visible wavelengths. A more
thorough analysis of the observations is being conducted to look for
small (order approximately 1%) brightness variations and set upper
limits on the brightness of the impact effects at visible wavelengths.

01.40-P Drossart P. Barucci A. Colas F. Crovisier J. Fulchignoni M.
Lecacheux J. Roos-Serote M. Roques F.
Observations of Jupiter at Pic-du-Midi During the SL9 Impacts

Observations of Jupiter have been conducted at Pic-du-Midi, first at the
2 meter Bernard Lyot telescope with the spectrometer MU-SICOS, working
in the red part of the visible spectrum between 540-880 nm at a spectral
resolving power of 38000, second on the 1 meter telescope with the CCD
imaging system Astriane, working in the visible and the near infrared
(J-H-K filters). The main results are:

(1) Several emission lines have been detected on impacts L and Ql in
spectra taken on July 19 at 22:42 UT (impact L) and on July 20 at 20:24
UT (impact Ql). The strongest lines are the Na doublet at 589.0 and
589.6 nm . Other detected lines, not present in Jupiter spectra before
impacts are due to Li (670.8 nm), K (766.5 nm), Ca (657.3 nm) and H
alpha (656.3 nm). These lines are observed in spectra less than 30 min
after impact and disappear totally in less than 1 hour after impact.

(2) CCD images show the evolution of the impact sites at several time
scales. Images in K filter show the complete sequence of the impacts, in
particular, impacts H, L and Ql. A flash appears first at the limb,
which is observed in all the bands (J-H-K). Then, a bright emission
feature appears in filter K, which should correspond to the raise of the
plume in the high atmosphere of Jupiter, slowly decreasing. In the long
term, the evolution of the dark cloud is followed in various filters in
the visible (including interferential methane filter). These images will
be reduced to provide the albedo of the dark clouds in standard
photometric filters.

01.41-P Gehrels T. Scotti J. V.
Spacewatch Observations and Astrometry of P/Shoemaker-Levy 9

Observations were made up to the time of impact, mostly with the
Spacewatch Telescope which is the 0.9-m Newtonian Reflector of the
Steward Observatory on Kitt Peak. The astrometry was used for the orbit
and impact predictions. Size estimates are being updated; ours are
smaller than generally used in the modelling.

01.42-P Lambert J. V. Africano J. L. Nishimoto D. L. Nolan R. A.
Kervin P. Medrano R.
Five Micron Observations of the Comet Shoemaker-Levy 9 Jupiter
C, F, and R Impacts from the Air Force Maui Optical Station

Near infrared, 4.8 - 5.05 micron, imagery of the C, F, and R fragment
impacts of Comet Shoemaker-Levy 9 on Jupiter was collected using the
Enhanced Longwave Spectrometer/Imager (ELSI) on the 1.6-meter AMOS
telescope at the Air Force Maui Optical Station. The impact sites were
visible only after they reached the planet's limb, then faded rapidly,
within ten minutes, leaving persistent spots slightly brighter than the
surrounding area. These spots appeared to lag the corresponding dark
impact spots observed in visible imagery recorded using the 1.2-meter
MOTIF telescope by ten to fifteen degrees in rotation. Extensive visible
and infrared observations were also made during the periods between
observable impacts. Several northern hemisphere spots and other
structures not apparent in the pre-impact five micron imagery were
observed. Five micron light curves for the impacts and infrared and
visible imagery will be presented. A relative intensity light curve for
the R-fragment impact is plotted and available in hard copy.

01.43-P Olson K. M. Mumma M. J.
Simulations of the Breakup and Dynamical Evolution of Comet
Shoemaker/Levy 9 Employing a Swarm Model

We investigated the dynamical evolution of a swarm of fragments produced
by tidal disruption of the nucleus of comet Shoemaker-Levy 9, for
comparison with observed properties of the nuclear train. We take the
initial state of Ihe comet nucleus to be a loose agglomeration of
macroscopic cometesimals, each bound more strongly internally than to
each other. Tidal disruption of the nucleus near perijove then produces
an initial swarm of these cometesimals (approximately 10^6 particles),
with possibly one sizable core remaining. We performed gravitational N-
body simulations of the subsequent dynamical evolution of this initial
swarm, and explicitly accounted for collisions between swarm particles,
gravitational interactions among them, and interaction with the tidal
field of Jupiter. We used an N-body tree code designed for highly
parallel computers, and ran a large number of simulations with particle
numbers ranging from 16,384 up to 131,072. The volume filling factor of
the pre-split nucleus was allowed to vary from 0.001-1.0.

We find that the initial swarm is tidally stretched into an elongated
structure which grows in length as the simulation proceeds. We also find
that in ALL cases the swarm becomes gravitationally unstable and a
number of clumps of bodies form. The time at which the instability sets
in (relative to periJove), the number of clumps which form, and the
nature of the clumps depend on assumptions about the detailed
collisional process between bodies. When the pre-split nucleus is
optimally packed (filling factor = 1) and the colliding particles are
treated as hard spheres, the instability does not set in until 78 hours
(approximately 41 Rj) after periJove. When the radius of each body is
0.5 of that in the optimally packed case, and some dissipation is added
to the collisions, the initial swarm of particles is very unstable and
distinct clumps form very soon after periJove (approximately 9 hrs,
approximately 7.6 Rj). We have found no cases to date which were not
unstable to some degree at some point in the simulation.

In general, we find that a collection of gravitationally and
collisionally interacting particles is gravitationally unstable as it
passes through the tidal field of Jupiter for a large range of the
parameters considered by our model. This tends to support the theory
that comet Shoemaker-Levy 9 was initially fragmented into a large number
(10^5 - 10^6) of small cometesimals and also has implications for the
structure of comets in general. We find it difficult to draw detailed
conclusions about the initial state of the comet before breakup owing to
the fact that clumps of particles form for a large range of the free
parameters of the model. These clumps are similar in number and spacing
to the observed 'string of pearls' observed prior to impact with
Jupiter. They explain in a simple way virtually all of the observed
properties demonstrated by the 'fragments' up to and including
disruption in the Jovian stratosphere.

This work was supported by the NASA Planetary Astronomy Program RTOP
196-41-54 and by the NASA/ESS HPCC program.

01.44-P Tozzi G. P. Richichi A. Ferrara A. Calamai G.
IR Photometric Observations of the Impact of Comet SL9 with

The impacts of fragments A, H, Q 1, Q2, T, and U of the comet Shoemaker-
Levy 9 (1993e) with Jupiter have been observed in the IR with the 2.2m
telescope of the Calar Alto Observatory. The sampling rate of the fast
photometer was typically 100-500 Hz, thus allowing us to follow with
high time resolution the evolution of the phenomena. The filters used
were a broad band K (lambda(sub)0 = 2.20 microns, delta lambda = 0.47
micron), and a narrow band ice filter (lambda(sub)0 = 3.12 micron, delta
lambda = 0.14 micron). The diaphragm was in all cases 15 arcsec.

During impact A, we monitored the luminosity of Europa at K. No evidence
of significant variations were revealed, within a dynamic range of 6-7
mag of the satellite's brightness (0.4%-0.1%). During all other events,
we monitored directly Jupiter's disk, centering at the expected impact
point. While no evidence of enhanced emission was revealed for the T and
U impacts, the H and Q impacts gave origin to increases in luminosity
which at the peak exceeded the emission of the planet's surface (in our
diaphragm) by a factor of approximately 25 - 50.

We present the corresponding lightcurves, which extend over about 0.5-1
hr with a (rebinned) resolution of 1 s. It is noteworthy that in both
cases a small "precursor" event is seen, at approximately 1-2 minutes
after the predicted and measured impact time and about 10 minutes before
the actual development of the fireball. It is also interesting to note
that while the Q1 fireball was extremely bright, the Q2 fireball
apparently did not produce significant emission. Another interesting
feature is the presence of a plateau in the decreasing part of the
lightcurve, especially pronounced in the case of the H fireball.
Comparison with other lightcurves taken at different wavelengths will be
shown and possible interpretation will be given.

01.45-P Wiedemann G. Encrenaz Th. Schultz R. Stuewe J. A.
Drossart P. Crovisier J.
Near-Infrared Spectroscopy of Jupiter at the Time of SL9 Impact

Near-IR spectra of Jupiter have been recorded with the IRSPEC imaging
spectrometer at the 3.5m NTT (ESO, La Silla), with the 4.4 arcsec slit
aligned along the parallel of impacts (latitude l= -44 degrees). After
the last impact, spectra were also recorded in the Northern hemisphere
at l=+44 degrees (see Schulz et al, this issue). Spectra were recorded
around the H2 S(1) quadrupole line at 2.12 microns, and around the 3.533
micron H(sub)^+ line.

Most of the impact sites (and later their images at l=+44 degrees) were
easily detected at 2.12 microns. Shortly after the impacts, spectra
recorded in the 2.107-2.135 micron range with a resolving power of 2000
show a drastic increase of the signal and a change in slope, with a
maximum peaking toward shorter wavelengths. The simplest explanation is
that there is a very strong scattering over a newly formed stratospheric
haze. This strong component sometimes hides the H2 quadrupole line,
which is detected when the continuum is low enough. These data are being
analysed to derive information about the particle size and the
stratospheric temperature at pressure levels of about 0.1-1 microbar.

Emission lines of H(sub)3^+ were detected in the 3.501-3.566 micron
range, with a resolving power of 1700, over many impact sites. As a
general rule, these emissions were stronger over more evolved sites, as
compared to the fresh impact sites. In the case of strongest emissions,
the multiplet of H(sub)3^+ previously detected in auroral regions was
observed, which will allow the determination of both the H(sub)3^+
column density and the stratospheric temperature around 10-100 nanobars.

01.46-P Xie X. Olson K. Mumma M. J.
Constraints on the Masses and Sizes of SL9 Fragments

We have performed N-body simulations of the infall of the fragments of
comet Shoemaker/Levy 9 from apoJove to shortly before impact with
Jupiter. These simulations assumed that each fragment is itself composed
of a large number of small bodies (the swarm model). It was also assumed
that the bodies interact solely through their self gravitational
interactions and also with the tidal field of Jupiter. Collisions
between the bodies have been ignored for this set of simulations. The
swarm of bodies is assumed to be initially in vireal equilibrium at
apoJove and then follows an orbit which will bring it into collision
with Jupiter. For this study we have varied the swarm mass (10^l3 -
10^l5 gms) as well as its radii (10 -100 km) to determine for what range
of these parameters the swarms hold together and would impact the
surface of Jupiter as a more or less solid object. The simulations show
that for the range of masses considered here and a critical size of
approximately 10 km, the swarms remain bound up until a few days before

01.47-P Reitsema H. J. Hubbard W. B. Roques F. Peyrot A. Vignand M.
Coordinated Imaging of SL-9 Impact Events: Observations from
Les Makes, La Reunion

High time resolution photometric observations of impact events and
imaging of impact sites were obtained from four globally-distributed
distributed sites using similar CCD systems. We report here data
obtained from Ile de La Reunion (at latitude -21:11:57, longitude
E55:24:35, elevation 995 m) which complete the observing net which also
included sites in Hawaii, Chile and Australia. Data were obtained each
night from 15 July to 24 July with the 36 cm telescope of the
Observatoire des Makes and seven filters covering B, V, R, and two
methane bands and corresponding continuum bands. Impact events A, E, H,
and S were observed with 0.4-sec time resolution during the impact
times, while slower multi-wavelength image sequences were taken during
impact events P, Q2, and T. Rapid image sequences of events A, E, and S
were obtained in the methane 890 nm filter, while the rapid image
sequence of event H was obtained using the B filter. However ongoing
analyses of impact events have yet to show any discrete signal from any
of these atmospheric entry events. Dark atmospheric disturbances and
occasional bright methane features were imaged at impact locations. The
poster will display representative images and data. Supported by NASA
Grants NAGW-1555 and NAGW-4072.

01.48-P Max C. Gavel D. Johansson E. Bissinger H. An J.
Lawrence T. Olivier S. Liu M. Bradford W.
Impacts of Comet Shoemaker-Levy: Speckle Imaging in the Visible
Using the 3 Meter Telescope at Lick Observatory

We performed speckle imaging of the impacts of Comet Shoemaker-Levy 9
with Jupiter at the Lick Observatory 3 meter telescope, from July 19
through 22, 1994. Images were obtained at wavelengths of 0.55, 0.7, and
0.85 microns. Using exposure times of 200 - 300 msec, we are able to
obtain high resolution images using a bispectral speckle reconstruction
algorithm from 20 to 40 individual speckle frames. We were able to
obtain continuous data over periods of 3 to 4 hours for several of the
impact points, and have made time-lapse movies following the rotation
and evolution of the impact spots. Features such as the dark "eyebrow"
and the interior darker region of the impact points are clearly
distinguished. In this presentation we shall show both images and
videos of our speckle reconstructions, and shall present preliminary
conclusions concerning the evolution of impact point structure over
periods of hours to days.

This research was supported by the US Department of Energy under
contract No. E-7405-ENG-48 to the University of California, and by the
Institute of Geophysics and Planetary Physics at LLNL.

01.49-P Lopez-Moreno J. J. Rodriguez J. Formisano V.
CCD Images of the Shoemaker-Levy 9 Impacts on Jupiter from
Sierra Nevada Observatory

A continuous coverage with CCD observations has been performed by using
the 1.52 telescope of the Sierra Nevada Observatory in Granada, Spain,
during the week (16-22 July) of the SL9 impacts. The telescope was
equipped with a direct CCD with filters and with a multispectral imaging
system (VNIR) covering from 350 to 1005 nm. During this period a daily
average of 50 CCD images during the 3 hours of visibility of Jupiter
were taken. We have also made observations of the impact regions some
days after the collisions (26, 27 and 29 July) in which the remaining of
the impacts were still clearly visible. The CCD was equipped with 6
filters covering the visible spectrum from 426 to 829 nm.

We present here some results on the evolution of the spot H from its
first appearance (H0) in July 18 to its seven passage (H7) on July 21.
We present ratio of the images at ammonia (648 nm) and methane (727 nm)
bands to the one obtained in the continuum at 635 nm and make
estimations of the depth of the features for each appearance.

This work was supported by CICYT under grants ESP-93-0338 and ESP-1290-

SESSION 02 ....... Comet Shoemaker-Levy 9 II
Monday, 10:30 - 12:00 Crystal Ballroom
Reta Beebe and William B. Hubbard, Moderators

02.01 Noll K. S.* McGrath M. Weaver H. A. Atreya S. Caldwell J. J.
Trafton L. Yelle R.
Hubble Space Telescope Spectroscopic Observations of Jupiter
After the Impact of Comet SL9

Dramatic changes in Jupiter's UV spectrum were recorded with Hubble
Space Telescope's (HST) Faint Object Spectrograph (FOS) and High
Resolution Spectrograph (HRS) following the impact of fragment G.
Darkening at all wavelengths and absorption and emission from a wide
variety of molecules and atoms were observed. In particular four
different sulfur-containing molecules were identified in quantities that
indicate the major source of sulfur is almost certainly Jupiter. This
observation provides a constraint on the depth of penetration of the
fragments and is also the first observation of sulfur in Jupiter.

A ratio of an 18 July FOS spectrum of the G impact site 3 hours after
impact to one obtained on 14 July effectively removes the solar spectrum
and normal jovian absorptions. The ratio shows strong absorption bands
of S2, CS2, and NH3; evidence for H2S is also present in the shape of
the absorption near 210 nm. The estimated column abundance of S2, N
about l0^l9 cm^-2 corresponds to a mass of sulfur of m similar to 10l4g
in our 0.9 arcsec field of view. This lower limit is a significantly
greater mass of sulfur than could be present in a l km diameter
impactor. The second post-impact FOS spectrum was obtained on 21 July,
approximately 30 minutes after the impact of fragment S. In this
spectrum we identify all the molecules identified in the 18 July
spectrum. In addition we observe strong, narrow emission lines from
MgII, MgI, SiI, FeI, and CS. Lower limits for the abundances of the
metal atoms are about l part in 10^6 of a l0^14 g impactor, easily
consistent with expected abundances of these elements in the impactors.
The remaining two FOS spectra obtained in August, on 9 August and 23
August, show an increase in NH3 relative to the CS2 absorption while the
S2 is no longer detectable.

We searched for Si- and 0- containing molecules in both EIRS and FOS
spectra. Derived upper limits indicate that very little of either
element is present in gas-phase species after the impacts. This evidence
may constrain the initial elemental abundances of the impactor, i.e. SL9
may have been a low-volatile object.

This research is supported by grant GO-5624.12-93A from the Space
Telescope Science Institute. Significant contributions to this work were
made by Paul Feldman, Chris Barnet, Diane Gilmore, Scott Edgington, Ed
Smith, and by Alex Storrs, Andy Lubenow, David Taylor, and many others
at STScI.

02.02 Yelle R. V. McGrath M. Noll K. Weaver H. Atreya S.
Caldwell J. Trafton L.
Spectroscopic Constraints on the Evolution of Sulfur-Bearing
Molecules in the Jovian Atmosphere

Ultraviolet spectroscopic observations of Jupiter's atmosphere
subsequent to the impact of comet p/Shoemaker-Levy 9 (SL9) have detected
the sulfur-bearing molecules S2, CS2, CS, and H2S. Several observations
were carried out with the Hubble Space Telescope Faint Object
Spectrograph (FOS) in the region from 170 to 330 nm with a spectral
resolution of 0.4 nm. A spectrum from the G impact site obtained on 18
July 1994 has as its most prominent feature a series of absorption bands
from 250-300 nm which are clearly due to the S2 (B-X) transition. In the
190-220 nm region this spectrum also contains absorption features due to
CS2 as well as absorption features attributed to NH3 and H2S. Spectra
obtained under similar viewing conditions on 9 and 24 August 1994
continue to show the NH3, CS2, and H2S absorption features but the S2
absorption is greatly reduced or absent. Spectra acquired from Jupiter's
limb on 21 July 1994 show the absorption features mentioned above but
also contain emission features due to CS and several metallic species.
Noll et al. (this session) will present an overview of FOS results and
implications. This presentation will focus on the ramifications of FOS
spectroscopy on the post-impact evolution of the sulfur-bearing species.
The observed changes in the spectra may be due to spatial inhomgeneities
or temporal changes. In this talk we will explore the latter
possibility. The S2 bands are the sulfur analogs of the O2 Schumann-
Runge band system which plays a critical role in terrestrial atmospheric
chemistry. Absorption in these S2 bands leads to production of atomic
sulfur, which will react with ambient atmospheric species further
altering the composition of the atmosphere.

This research is supported by grant GO-5624.12-93A from the Space
Telescope Science Institute. Significant contributions to this work were
made by Paul Feldman, Chris Barnet, Diane Gilmore, Scott Edgington, and
Ed Smith, and by Alex Storrs, Andy Lubenow, David Taylor, and many
others at STScI.

02.03 Lellouch E.* Paubert G. Moreno R. Festou M.
Bockelee-Morvan D. Colom P. Crovisier J. Encrenaz T.
Gautier D. Marten A. Despois D. Strobel D.
Millimeter-Wave Observations of the Jupiter/Comet
Shoemaker-Levy 9 Collision from IRAM 30-m Telescope: CO, CS,
and OCS

Millimeter-wave observations of the collision of comet Shoemaker-Levy 9
with Jupiter were conducted at the IRAM 30-m telescope on July 17-28,
1994. These observations led to the detection of emission lines of CO
at 230.538 GHz and CS at 244.935 GHz on a number of impact sites
(C,E,G,H,K,L,Q+G for CO; K,L,Q+G,W for CS). The typical linewidth (FWHM)
is 2 MHz, suggesting a primarily collisional broadening (with a probable
contribution due to velocity smearing within the impact sites) and
indicating a formation level of about 1 mbar. For a few sites, the lines
were monitored over several days, providing information on their
temporal evolution. Generally, the lines appear to be weak immediately
after impact (0.1-0.2 K antenna temperature contrast), then to increase
to reach a maximum (0.5-1 K) one or two days after impact, then to
decrease. In particular, on impact site Q+G, the decay of the emissions
from July 21 to July 23 and 25 was observed. On July 28, the CO and CS
lines were detected in absorption, with perhaps an emission core in the
case of CO. All these variations probably result from a combination of
variations in (i) the horizontal extent of the sites (ii) the vertical
distribution of the molecules and (iii) the temperature profile. We also
report the probable detection of OCS at 218.903 GHz on impact site W on
July 22, and upper limits on CO^+, CH3OH, HC3N, H2S, SO2, CN, and H2CO.
Very preliminary modelling, assuming entire deposition/formation of the
molecules near 1 mbar, suggests that the strongest CO and CS lines
require approximately 10^14g of CO and 5 x 10^11g of CS. On impact site
W, the lines indicate about 1.5 x 10^11g of CS and 1.5 x 10^12g of OCS.
The corresponding relative abundances (in volume) for CS:OCS:CO are
about 1:7:350. A cometary origin for the O and S seen in CO, CS, and OCS
seems therefore plausible at the present stage.

02.04 Encrenaz Th.* Schulz R. Stuewe J. A. Wiedemann G.
Drossart P. Crovisier J.
Detection of 3.5 Micron CH4 emission at the Time of SL9-H
Impact: An Estimate of Rotational Temperatures

We have monitored Jupiter in the 3.501-3.566 micron range shortly
after the impact of fragment H, using the IRSPEC imaging spectrometer
at the 3.5m NTT (ESO, La Silla). The resolving power was 1700; the
4.4 arcsec slit was aligned along the parallel of impacts, and the
pixel size along this axis was 2.2 arcsec. Our observations started
on July 18, UT 19:46, i.e. 13 mn after the impact, at a rate of 1
image per minute. A very strong emission was detected over the whole
spectral range, due to high J-value multiplets of the CH4 nu(sub)3
band (J=14 to J=18). The signal intensity decreased exponentially
with a timescale of about 5 minutes, and was detectable for about 30

At the beginning of the sequence, the CH4 emission extends over about
10 arcsec along the slit. The slope of the emission shows drastic
variations over the emission regions, corresponding to a large range
of rotational temperatures. Assuming, as a first step, that methane
lines are unsaturated, the rotational temperature, in the first
image, is estimated to about 700K at the center (emission peak). On
the edge (leading side), a different spectrum is observed, with a
much weaker intensity, which may indicate different temperature
profiles in the center and on the edge.

A modeling of these emissions is in progress, which will give
important constraints on the temperature profiles at the impact site
and around.

02.05 Friedson A. J.* Goguen J. D. Orton G. S. Hoffmann W. F.
Dayal A. Deutsch L. K. Wells W. K. Hora J. L. Spitale J. N.
Fazio G. G.
Thermal Infrared Observations of the Impact of Fragment R at
the IRTF

The impact of fragment R was observed at thermal infrared wave-lengths
of 7.8, 10.3 and 12.2 micrometers from the NASA/Infrared Telescope
Facility on July 21 (UT), using the MIRAC2 mid-infrared array camera
(Hoffmann et al., Infrared Phys. Technol. 35, 175, 1993). Thermal
emission at the three wavelengths was sampled sequentially using a 2%
circular variable filter, with an average time interval of 17 seconds
between observations at different wavelengths. Continuous imaging of
Jupiter in this mode began at 5:08 UT and extended to 5:55 UT. Clear
evidence for enhanced emission from the impact region first appears at
5:41 UT, with the peak in fireball emission at all three wavelengths
occurring 3.5 minutes later. The excess flux at each wavelength then
falls to less than 20% of the peak value over a period of about 4-5
minutes. Results based on preliminary calibrations of the 7.8 and 10.3
micrometer excess flux indicate a source color temperature of 3650
+ll00/-600 K and, on the assumption that the source was optically thick,
a solid angle of 1.5 +/- 0.5x1O^-l4 sr (equivalent to that subtended by
a sphere about 110 km in diameter) at the time of peak emission. Fifty
seconds later, the color temperature has fallen to 2050 +l50/-200 K and
the solid angle is 3.2 +/- 0.6x1O^-l4 sr (equivalent spherical diameter
about 160 km). Work in progress includes i) refining the absolute
calibrations, particularly for the 12.2 micrometer data; ii) including
the calibrated 12.2 micrometer data in the analysis to determine whether
the excess fluxes follow a black body spectrum; and iii) least-squares
fitting a physical model of the fireball to the data as a function of

02.06 Livengood T. A.* Bjoraker G. L. Kostiuk T. Romani P. N.
Kaufl H. L. Wiedemann G. Ivchenko N. Mosser B. Sauvage M.
Thermal-Infrared Imaging of the SL9 Impact Sites Using TIMMI at
the European Southern Observatory

We present initial results from a successful program of thermal-IR
imaging that was conducted as part of the European Southern Observatory
SL9 observing campaign using the Thermal-Infrared Multimode Instrument
(TIMMI) on the 3.6m telescope, during and following the collision of
comet Shoemaker-Levy 9 with Jupiter. We obtained complete or nearly-
complete 10 micron light curves during the A and H fragment impacts
(3.26-second sampling), and extensive multiwavelength imaging of impact
sites after the initial explosive phenomena, with filters at 5,
7.8,10.1,10.65,12.1, and 13.3 microns. Observations extended up to seven
nights following the last impacts, with spatial sampling of 0.63 or 0.47
arcsec per pixel on a 64 x 64 pixel detector array. Many impact sites
were still discernible as of our last observations, possibly due to
either the persistence of non-equilibrium species (e.g., NH3), heating
by aerosols in the stratosphere, or simply long radiative time constants
for gaseous emission. The impact sites are sufficiently large that their
spatial extent is resolved in these images. These observations were
conducted jointly with a program of B. Mosser (related abstract, this

TAL acknowledges support as an NRC Research Associate at NASA/GSFC
during this work.

02.07 Halthore R.* Bjoraker G. Livengood T. Allen J. Jr.
Kostiuk T. Stolovy S. Herter T. Kaufl U.
Energy Deposited in the Jovian Stratosphere by the Impacts of
Comet Shoemaker-Levy 9

Light curves in the 7.7 micrometer band observed using the Kuiper
Echelle Grating Spectrometer (KEGS), for the G and K impact sites of
comet Shoemaker-Levy9 on Jupiter show a rapid cooling of the hot core in
the stratosphere. As the impact sites become visible over the limb,
temperatures of the hot cores derived from the water emission lines are
seen to be in the 500 K range (Bjoraker et al., this meeting). The
horizontal size of the hot core can be estimated using the methane
emission at 7.7 micrometers provided methane abundance is not altered in
the hot spot. From the light curves in the methane emission, the
temporal evolution of the size and temperature of the hot core are
obtained. Results will be compared with those of TIMMI camera (ESO), an
imaging instrument that observed the impacts in the 10 micrometer band.
TIMMI will allow us to study the long term evolution of the impact

02.08 Bjoraker G. L.* Herter T. Gull G. Stolovy S. Pirger B.
Detection of Water in the Fireball of Fragments G and K of
Comet Shoemaker-Levy 9

We observed Jupiter on 17-19 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 20 spatial and 128
spectral elements at a resolving power of 9000. The principal objectives
were to search for H20 emission lines near 22 and 24 micrometers and to
measure the stratospheric temperature using CH4 emission lines at 7.7
micrometers. It was necessary to switch between these 3 wavelengths
during each flight. At the peak of the fireball for the G and K
fragments, the intensity of the CH4 emission lines increased by a factor
of 30 from pre-crash levels. Within the 7.67-7 72 micron bandpass, we
detected 3 emission lines of H20. The lower state energy levels of these
transitions are populated only at high temperatures (about 5OO K). The
intensity of the water lines dropped by an order of magnitude in 15
minutes. Spectra at 22 and 24 micrometers were obtained after the peak
of each fireball. No dramatic change in either the continuum level or
depths of telluric H20 lines were observed. The long wavelength channels
are sensitive to the temperature of the upper troposphere and to large
amounts of modestly heated water. We conclude that the water came from
the impactor, not from Jupiter, based on a combination of KAO and
ground-based data. A quantitative estimate of the water abundance will
be required before determining whether the impactor was a comet or an

02.09 Sprague A. L.* Hunten D. M. Witteborn F. C. Kozlowski R. W. H.
Wooden D. H. Bjoraker G.
KAO Observations of Jupiter During and Following the Impact of
Comet SL-9 Fragments R and W Using HIFOGS (4.9-9.4 and 9.3-14.5

Observations of Jupiter during and following the impacts of two segments
of Comet Shoemaker-Levy-9 were made July 21 and July 22 1994 from the
Kuiper Airborne Observatory using the High Efficiency Infrared Faint
Object Grating Spectrograph (HIFOGS) [1]. An emission peak at 6.6
micrometers identified as H20 is prominent following the R impact and is
seen in spectra following W. Line ratios indicate that the water
temperature is near 500K. The Q branch of the 7.8 micrometer band of CH4
rapidly rose to 11 times brighter (5^h 45^m) than pre-impact
measurements. It decayed by 25% in the next 3 minutes (by 5^h 48^m)
following the R impact. A large increase was also seen at these
wavelengths following W but the factor is not yet known. Also seen is a
large emission increase of acetylene at 13.7 micrometers following both
impacts and other, unidentified emissions.

The KAO new out of Melbourne, Australia. HIFOGS has cryogenically cooled
aperture and filter wheels, grating, and a linear array of 120 Bi-doped
Si detectors. To increase the spectral range slightly, and to increase
the resolution (decrease the wavelength sampling increment), standard
stars and Jupiter were measured at four grating settings spanning
wavelengths of 4.9 to 9.4 micrometers and 9.3 to 14.5 micrometers.
Residual telluric absorptions are removed by calibrating to spectra of
alpha CMa and alpha Cen when possible or by using atmospheric
transmissions generated by the HITRAN data base and the NASA Ames ATRAN
program. Spectral shape is removed by correcting ratios using alpha CMa
which has been well-characterized using the KAO in previous flights.

1. Witteborn F.C. et al. Astro. Soc. Pacific Conf. Series Astro. IR
Spec. Conf. 365-372 (1993).
2. Cohen et al. Astro. J. 104, 5, 2030-2044 (1992).
3. Cohen et al. Astro. J. 104, 5, 2045-2052 (1992).

02.10-P Harrington J. Dowling T. E. Hammel H. B. Mills J. R.
Hoffmann W. F. Dayal A. Wells K. Sprague A. Hora J. L.
Deutsch L. K. Fazio G. G. Baines K. H. Friedson A. J.
Orton G. S. Yanamandra-Ficher P. A.
IRTF and HST Observations of the Dynamic Response of Jupiter's
Atmosphere to the Impact of Comet Shoemaker-Levy 9

We present images of Jupiter to be obtained immediately before and for
several weeks after fragments of comet Shoemaker-Levy 9 impact the
planet. We report dynamical effects of the impacts on the atmosphere. If
sufficient energy is deposited in the atmosphere, the predicted
behavior, including inertia-gravity wave excitation and vortex
generation, acts as an atmospheric probe that constrains structure and
dynamics. Imaging of reflected sunlight with the Hubble Space Telescope
(HST) Wide-Field and Planetary Camera 2 (2159, 2553, 3344, 4090, 5407,
8922, 9534 Angstrom) and NASA Infrared Telescope Facility (IRTF) NSFCAM
facility camera (1-3 micrometers) probes the upper ammonia cloud level
(approximately 150 mbar). Imaging of thermal emission with NSFCAM (4-5
micrometers) and the Arizona/SAO/NRL Mid-Infrared Array Camera (5-20
micrometers) on the IRTF probes the stratosphere and troposphere
(approximately 30-500 mbar) with a vertical resolution of 1-2 scale

02.11-P Zahnle K. Mac Low M.-M.
More Consequences of the Collision of a Comet and Jupiter

The pieces of the former Comet Shoemaker-Levy 9 struck Jupiter during
the week of July 16-22 1994. We divide the general impact phenomena into
four phases. (1) Atmospheric entry should have been accompanied by an
entry flash, probably only detectable by the Galileo spacecraft. (2)
Aerodynamic forces cause the impactor to explode. We show that the
apparent altitude of the explosions is consistent with the analytical
model proposed by Zahnle and Chyba et al, with numerical models
calculated by Mac Low and Zahnle and Yabe et al., and with dynamically-
based estimates of fragment masses suggested by Scotti and Melosh and
Asphaug and Benz. (3) The explosion produces a fireball that propagates
most quickly up the wake, as predicted by Takata et al. and Crawford et
al. The fireball's luminosity fades over a timescale of 30-60 seconds.
(4) Each explosion produced a plume that reached thousands of kilometers
above the cloudtops. The plumes appear to have been dusty and so
illuminated when they reached sunlight. The plumes rose and fell over a
period of ten or twenty minutes, governed by the simple laws of
ballistics. When plume material (dust) fell back on the atmosphere it
was reheated and glowed brightly in the thermal infrared for a period of
order ten to twenty minutes; a large fraction of the impact energy was
spent thus. Geometry introduces additional factors that complicate
interpretation of the observed light curves.

02.12-P De Jong E. M. Ingersoll A. P. Toigo A.
Video Representation of Shoemaker-Levy 9 Impact

We present a video animation of the five phases: entry, plume, pressure
wave, internal gravity wave, and long-lived vortex following the impact
of comet Shoemaker-Levy 9 into Jupiter. The entry phase lasts a few
seconds and ends when the bolide has lost its initial kinetic energy.
The plume (fireball) phase lasts a few minutes and ends when the ejecta
have fallen back and the strong shock waves have dissipated. The
pressure wave phase lasts for 1-2 hours and ends when the waves'
amplitude drops to unobservable levels. The gravity wave phase lasts for
1-2 days and ends when the wavefronts have been broken up by the zonal
winds and the waves' amplitudes have fallen to unobservable levels. The
vortex phase may last indefinitely if the vortex can adjust to a stable
configuration in the zonal shear flow. We have borrowed heavily from
numerous authors and, where necessary, have parameterized their results.
The objective in producing the video is educational -- to provide
physical insight into the various phenomena and their interrelations in
space and time.

References: T. Ahrens et al. GRL 21,1087 (1994), J. Harrington et al.
Nature 368, 525 (1994), D. Hunten et al. GRL 21, 1091 (1994), A.
Ingersoll et al. GRL 21, 1083, (1994), M. Marley Ap. J. Lett. 427, L63
(1994), K. Zahnle and M.-M. Mac Low Icarus 108, 1 (1994)

02.13-P Marley M. S. Dayal A. Deutsch L. K. Fazio G. G.
Hoffmann W. F. Hora J. L. Hunten D. M. Sprague A. L.
Sykes M. V. Walter C. Wells K. W.
A Search for Seismic Waves Launched by the Impact of Comet

The impact and explosion of each fragment of comet Shoemaker-Levy/9
(1993e) into Jupiter's atmosphere was predicted to launch an acoustic
"seismic" wave that would probe the interior structure of the planet.
We attempted to detect the thermal infrared signature of this wave as it
returned to the visible atmosphere after refraction in the interior of
Jupiter. To do so the mid-infrared array camera MIRAC2 was deployed on
the IRTF to search for perturbations in the 7.8 micrometer CH4
stratospheric emission feature following several impacts. The planned
S/N ratio of 100 per pixel in an exposure of 140 sec would have allowed
detection of the approximately 0.1K temperature fluctuations expected
after impacts exceeding approximately 10^28 erg. Owing to uncooperative
weather, the seismic wave search was conducted only in the aftermath of
the moderate-sized R impact and technical problems limited the exposure
time. Our dataset thus consists of 240 7.85 micrometer images of
Jupiter obtained in the 90 minutes after the impact. The seismic wave
was expected to appear as a growing ring centered on the impact site
during this time period. No such structure is obvious in our current,
slightly-processed images. With substantial further reduction and
analysis we expect to either detect the wave feature or place an upper
limit on its stratospheric temperature amplitude and, ultimately, the
impact energy.

02.14-P Lognonne Ph. Billebaud F. Vauglin I. Merlin Ph. Sybille F.
Mosser B. Lagage P. O. Gautier D. Drossart P.
Seismic Waves Generated by the SL-9 Impact

The 10 micron camera from the Observatoire de Lyon was mounted on CFHT
between July 21 and July 27 1994. Due to hurricane Emilia, no
observations were done July 21, e.g. for the last impacts V and W and
only post-impact data were obtained between July 23 and July 27. 5 hours
of continuous observations were recorded in the 10.57-12.94 micron
filter between July 24 and 27, at a rate of one image every 30 seconds,
for a search of trapped pressure waves with period greater than 5.5
minutes, and several images were taken with 4.5-5, 9.16-10.08 microns
and CVF filters for a search of slow thermal changes, related either to
gravity waves or non-reversible heating of the Jovian atmosphere due to
surface waves. The post impact seismic program on CFHT was coordinated
with two other telescopes, all using the same type of instrument: TIMMI
camera at the 3.6-m telescope at ESO (Chili), and CAMIRAS, from CEA, at
the NOT (Canary Islands).

The analysis of these data is under progress. Synthetic simulations on
the excitation of normal modes and surface wave have shown that the
post-impact seismic waves are observable for impacts with energy greater
than 10^{21} J. However, the seismic waves are strongly focused at the
antipode and a seismic signature at the antipode remains possible for
lower impact energies [Lognonne et al., 1994]. If positive, seismic
observation will allow the inversion of a profile of seismic velocity.
From the amplitude of seismic signals, it will also be possible to give
a range for the explosion released by the impact.

02.15-P Collins M. D. McDonald B. E. Kuperman W. A. Siegmann W. L.
Jovian Acoustics and Comet Shoemaker-Levy 9

A three-dimensional acoustic model that handles spatially varying
acoustic parameters and wind fields has been developed and applied to
model global-scale sound propagation from the impact sites of the
fragments of Comet Shoemaker-Levy 9. To achieve these modeling
capabilities, it was necessary to neglect nonlinear effects. Although
nonlinear effects are important near the impact sites, the spatial
distribution of compressional wave energy predicted by the linear model
should be qualitatively representative. The vertical dependence of the
acoustic parameters and the latitudinal dependence of the wind are known
from Voyager data. The Jovian atmosphere contains a sound channel, which
is a layer of relatively low sound speed that ducts acoustic energy by
refraction and contains the cloud layers. Since several of the fragments
exploded just above or within the cloud layers, a great deal of acoustic
energy was released in the sound channel. The zonal winds consist of
cells that blow in opposite directions and act as acoustic waveguides.
The acoustic model predicts that energy in the sound channel is pinched
into well-defined beams by the zonal winds. The two strongest beams form
in wind cells to the east and west of the impact sites and remain well
defined at great distances from the impact sites. Since these beams are
caused by geometric effects and delimited by acoustic shadow regions,
they should be the most prominent acoustic features away from the impact
sites. Images from the Hubble Space Telescope show evidence of
cylindrical acoustic wave fronts propagating in the sound channel near
the impact sites. Processed data might reveal structured wave fronts
away from the impact sites. This work was supported by ONR.

02.16-P Maillard J. P. Drossart P. Bezard B. de Bergh C. Lellouch E.
Martin A. Caldwell J. Owen T. Atreya S. Waite H. Yelle R.
FTS Spectroscopy of the SL9 Impact Sites on Jupiter

The collision of comet Shoemaker-Levy 1993e with Jupiter has been
observed with the Fourier Transform Spectrometer at the CFHT. A 2.5-
arcsec aperture (corresponding to 8000 km on Jupiter) has been selected,
and the impact sites were observed by offset guiding on a galilean
satellite. The goal of the observations was a search for modifications
of the jovian composition after the impacts, at tropospheric and
stratospheric levels. As a priority, the 4.7-micron window was observed
to search for modifications of CO, with a resolution up to 0.1 cm^-l. In
the same spectral range signatures from PH3 and CH3D are also present.
Impact sites B, C, F, G and L were observed in this spectral range,
showing increased cloud opacity but no dramatic changes in the spectral
absorptions. Two fragments (C and R) were monitored at the time of the
impact, directly observable from Hawaii, in a 3.3-micron filter covering
the 2850-3020 cm^-l range, with a resolution up to 0.19 cm^-l. This
region is dominated by the R-branch of the nu(sub)3 band of methane
which blocks all the solar reflected light in normal jovian spectra. For
both impacts an emission peak appeared at 3000 cm^-1 with further
structures between 2860 and 2950 cm^-1, lasting less than 20 min.
Preliminary simulations with a 500-K temperature profile down to the 10-
micronbar level and normal below, indicate that these structures could
possibly be due to methane emission. Additional observations were made
in the 2.2-micron window on fresh impact sites (G and H) allowing the
detection of a haze continuum and of the pressure-induced spectrum of

02.17-P Korsun P. P. Sizonenko Yu. V.
High-resolution Spectra of the Satellite Reflections During the
Comet Shoemaker-Levy 9 Impacts on Jupiter

Long-slit CCD spectra of Jupiter were obtained with the 2.6-m telescope
of Crimean Astrophysical Observatory, ZTSh, on 19-22 July, 1994. One set
of the observations was carried out at Nasmith focus and the wavelength
region covered 4600-10,250 angstroms at a resolution of 4.5 angstroms.
The other one was carried out at coude focus and the spectra extended
over various parts of the same spectral region at a resolution ranging
from 0.2 to 0.8 angstroms. The spectrometer slit was aligned along the
parallel of the impact to provide a spatial analysis in this direction.

Combined investigations of Nasmith spectra, both the spatial profiles
for different spectral regions and extracted spectra for different sites
of the crash latitude, show that the impact spots have some spectral

The main results of our investigations we can summarize as

All impact sites have low albedo with respect to undisturbed surface of
Jupiter in at least 4600-8000 angstroms wavelength region. No obvious
changes in the spots spectra shape are apparent, except for 4600-5700
angstroms region where albedo of the spots slightly increase to the

Observed spots showed considerably less methane absorption in the 8900-
angstrom band. Such effect was somewhat lower in the 1-micrometer
CH(sub)4 band from the H and D+G+S+R sites spectra and no detected in
the other observed methane bands.

We are grateful to R. E. Hershberg, Yu. F. Mal'kov, and V. I. Pronik
(CrAO) for their support and assistance in our observations.

02.18-P Wannier P. G. Spilker T. R.
Submillimeter Observations of the SL-9 Impact from KAO

We will use a heterodyne submillimeter radio receiver on NASA's Kuiper
Airborne Observatory to measure the abundance and vertical distribution
of water in the Jovian stratosphere immediately following the collision
of comet Shoemaker-Levy. Our specific target is the 548 GHz line of ^18O
water, minimizing interference by terrestrial water. The Jovian
stratosphere is normally dry but water vapor, once introduced, will
persist for relatively long periods, possibly years. Water may be
injected by the fireball following a deep penetration, or by sublimation
of cometary ice. After injection, lateral stratospheric circulation
should distribute the water over the planetary disk within a few days.
The heterodyne technique yields very high resolution (upsilon/delta
upsilon approximately equal to 10^6) spectra that can be used to infer
both depths of the observed water, from pressure broadening, and
longitudinal distributions, from timing and Doppler shifts.

02.19-P Orton G. Baines K. Friedson J. Goguen J.
Yanamandra-Fisher P. A'Hearn M. Esterle P. Lisse C.
Weaver H. Wellnitz D. Hoffmann W. Hunten D. Dayal A.
Sprague A. Sykes M. Wells K. Marley M. Fazio G. Deutsch L.
Hora J. Jewitt D. Joseph R. Deming D. Kostiuk T.
Bjoraker G. Fast K. Livengood T. Zipoy D. Griffith C.
Dowling T.
The NASA/IRTF SL9 Observing Campaign

The NASA Infrared Telescope Facility campaign (July 12 - August 7, 1994,
supplemented by later observations) used near- and middle-infrared
cameras NSFCAM and MIRAC2, near- and middle-infrared spectrometers
CSHELL and IRSHELL, and the Goddard heterodyne spectrometer in several
investigations. (1) Fragment K was observed an hour before its impact.
(2) The impact fireballs were investigated from fragments C (2.248
microns), G (2.295 and 4.780 microns) and R (7.85, 10.30 and 12.20
microns). (3) Particulates high in the stratosphere at impact lasted for
many weeks. (4) Temperatures were perturbed between 10 and 400 mbar
pressure. (5) The temperature field was searched for seismic waves, and
both the temperature and cloud field were searched for inertia-gravity
waves. (6) The north polar H(sub)3^+ aurora brightened after the
impacts; the stratospheric hot spot was suppressed during the impacts.
(7) NH3 was transported into the stratosphere, C2H4 and C2H6 abundances
was enhanced, and tropospheric PH3 was depleted. (8) Little influence of
the increased dust environment on Jupiter's ring was detected in the
near infrared.

02.20-P Smythe W. D. Carlson R. W. Weissman P. R. Hui J.
Segura M. E. Baines K. H. Matson D. L. Johnson T. V.
Leader F. E. Taylor F. W. Encrenaz T. Drossart P.
Galileo Infrared Observations of the Shoemaker/Levy-9 Impacts
of Jupiter

A direct view of the impact sites of the comet fragments from the
Galileo spacecraft allows a study of the fireball expected to occur
during the first minutes following each impact. Predicted temperatures
of a few thousand degrees place the spectral emission within the
wavelength range (0.7 to 5.2 microns) of the Near Infrared Mapping
Spectrometer, which will observe several of the events in seventeen
wavelength bands. The spectral bands chosen exhibit differing gas
opacities and allow sounding at various depths in the atmosphere. These
spectral positions include the Jovian spectral windows (e.g. 2.7 and 5
micron windows and shorter wavelength continuum regions) and a band for
H(sub)3^+ emission. To ensure observation of the impacts, given
spacecraft pointing errors, the scan platform will dither across Jupiter
with a 5 second period. Data will be recorded for 64 minutes for 4
impacts, although only a portion of these data will be selected for
playback. Initial pecks at the data for the G and R impacts show that
the scan platform and the NIMS instrument operated correctly. Normal
Jupiter spectra were obtained. Approximately the first 3 minutes of the
G impact spectra are expected to be available in time for the DPS
meeting. Preliminary results of the spectral properties and time
development of an observed fireball will be presented.

02.21-P Kim S. Dumas C. Orton G. Brooke T. Spencer J.
Spectroscopic Study of the SL9 Impact Areas of Jupiter with the

We obtained spectra of Jupiter's atmosphere in the 3 - 5 micron region
with the newly modified Infrared Spectrometer (IRS) on the 1.5-m
telescope at the Cerro Tololo Inter-American Observatory (CTIO) between
July 16 and 25, 1994 (UT), during and after the impacts of fragments of
comet Shoemaker-Levy 9. Analyzing the spectra of the impact sites, we
have found the following: (1) The upsilon(sub)3 band of CH4 has been
shown in emission on the major impact areas; (2) The spectral
characteristics of the impact clouds are very similar to that of the
usual polar haze. The polar haze is dark in UV and visible ranges, and
bright in methane bands in the visible range. The polar haze is bright
between 1.5 and 4 microns, wherever CH4 absorption is strong. The reason
why the impact areas are bright in the CH4 bands is that the impact
cloud altitude is high (approximately 1 mbar as we reported previously.
See DePoy et al., this conference), so that there is less CH4 absorption
above the impact clouds than above the normal cloud layers; (3) We found
brighter global H(sub)3^+ emissions than the normal condition. In
contrast, we found no significant increase in the H(sub)3^+ emissions at
the major impact sites, and frequently we did not detect H(sub)3^+
emissions finding only CH4 upsilon(sub)3 band emissions at the major
impact sites. We are now investigating the following possible phenomena:
( 1 ) CH4 upsilon(sub)3 band emissions at other latitudes, (2) altered
vertical and hortzontal distributions of atmospheric molecules, haze
particles, and cloud levels; (3) the presence of new molecules and
comparisons with model calculations; and (4) vertical mixing ratios and
temperatures of tropospheric molecules to determine whether fragment
penetration is deep enough to cause tropospheric mixing. As a part of a
network of internationally coordinated observations, including other
near-infrared spectroscopic obsetvations in the U.S. southwest, Hawaii
and Australia we are comparing our observations with other observers'
results to shape a unified scenario of the impacts.

02.22-P Dinelli B. Miller S. Achilleos N. Lam H. Tennyson J.
Jagod M.-F. Oka T. Geballe T. Brooke T. Ballester G.
Trafton L.
Infrared Spectroscopic Studies of the Impacts of Fragments B
and C of Comet Shoemaker Levy-9

The effects of the Fragment B and C impacts of SL9 on Jupiter were
monitored at wavelengths around 3.5 microns, sensitive to ionospheric
H(sub)3^+ emission, using the CGS4 long-slit spectrometer on UKIRT. Data
obtained prior to the impacts, with the slit aligned east-east along the
S45 region, showed that emission from the western region of the planet
was somewhat stronger than that from the east. But this pattern was
reversed with the approach and impact of Fragment B, even though this
fragment made hardly any impact deeper into Jupiter's atmosphere.

The Fragment C impact gave rise to emission due to high-J lines of
methane (principally P(18) and P(l9) of upsilon(sub)3), at intensities
which were at least two orders of magnitude greater than the ambient
ionospheric H(sub)3^+ emission. The most intense H(sub)3^+ line
(upsilon(sub)2 R(3,3)) in our spectra also appeared to brighten by this
amount. But, since the CH4/contiupsiloupsilonm emission swamped the
weaker lines we were monitoring, we cannot be sure that all this
intensity is due to H(sub)3^+. Initial fitting of CH4 lines to our
spectra suggests that temperatures in the plume some 15 miupsilontes
after nominal impact were around 1300K, falling to about half this
during the next 40 miupsilontes. But CH4 fits the spectra we obtained
poorly, indicating that other -- thus far unidentified -- species may be

02.23-P Crawford D. A. Boslough M. B. Robinson A. C. Trucano T. G.
Numerical Simulations of Fireball Growth and Material Motion
During Comet Shoemaker-Levy 9 Impact on Jupiter

The CTH shock-physlcs code ls used to study two- and three-dimenslonal
representations of the impacts of Comet Shoemaker-Levy 9 on Jupiter. The
simulations are divided into two closely linked parts. The first part
consists of two- and three-dimensional simulations of entry, deformation
and breakup of the cometary fragments. Realistic material models
incorporating melting, vaporization, dissociation and ionization of the
comet fragments and Jovian atmosphere are used. The comet fragments are
modeled as 1-, 2- and 3-km spheres of fully dense (0.95 g/cc) and porous
(0.3 g/cc) water ice, porous silicate (1.0 g/cc) and distended clouds of
unbound water and silicate fragments (average density 0.01-0.2 g/cc).
Depth of penetration and energy deposition per unit altitude are
tabulated for comparison with other models.

The second part of the simulation takes, as its input, material
location, velocity, pressure,,temperature and density from re-entry,
deformation and breakup studies. These are mapped into a three-
dimensional computational mesh for rendition of fireball growth and
evolution. Lagrangian tracer particles track the motion of cometary
debris and Jovian atmospheric constituents. At 120 seconds after the
impact of a fully dense 3-km cometary ice fragment (M = 1.4 x 10^16 g),
the total mass of the resulting fireball above the 1-bar level is 6.8 X
10^l6 g (4.9M) with 0.2% (1.5 x 10^14 g, 0.01M) of the fireball as
impactor material. At this time, 4.4 x 10^l6 g (3.1M) of atmosphere has
been uplifted above the l-bar level from below the tops of the Jovian
water clouds. Assuming a 10^-3 relative abundance of water in the water
clouds, our simulation shows that approximately 4 x 10^13 g of water
(0.003M) has been entrained in the lower portion of the fireball.
Dependence of material motion over the range of fragment models
described above will be discussed.

This work is supported by the National Science Foundation under
Agreement No. 9322118 and performed at Sandia National Laboratories
supported by the U.S. Department of Energy under contract DE-AC04-

02.24-P Biraud Y. G. Cuisenier M. Marten A. Rosenqvist J. Moreau D.
Muller C. Balega I. Chuntonov G. Maslov I.
Near Infrared Photometric Observations of the Comet SL9-Jupiter
Collision at the Zelenchuk Observatory

Near infrared observations of the comet impacts with Jupiter have been
performed at the SAO 6-m telescope (Zelenchuk/Caucasus) using a Fourier
spectrometer. Because of an unpredictable technical problem, the
observations were limited to spectro-photometry between 3.3 and 4.2
microns with a resolving power around 300. Observations were performed
on July 17, 19, 22, and 27, consisting of a total integration time of 3
hours. The K and W impacts were mainly monitored. Preliminary results
did not show any obvious variability of the 3-4 microns low resolution
Jovian spectra but a more detailed investigation of these data will be
reported in order to determine possible _small_ variations of the
Jupiter'continuum. In addition to infrared spectra, CCD cameras provided
visible information on the location, shape, and size of the impacts.
After filtering and digitization, a spatial resolution of 0.4 arcsecond
could be attained.

02.25-T Trafton L. M. Atreya S. K. Noll K. S. McGrath M. A.
Weaver H. F. Caldwell J. J. Yelle R. V.
Comet SL9-Jupiter Impact Events: HST Search for Enhanced
Stratospheric CO and H2S

During the SL9-Jupiter impact events, HST will obtain exploratory
spectra of the impact sites. One objective is to detect and measure
cometary and photochemically enhanced CO and H(sub)2S in Jupiter's
stratosphere. This will be attempted using the GHRS to obtain spectra of
the 4th Positive bands of CO in emission and the FOS to get spectra of
the H2S absorption bands from 2000 - 2700 Angstrom. The potential
enhancement of CO from these sources is significant and diagnostic of
the cometary CO + H20 content, and of the photochemistry between the
Jovian hydrocarbons and uplifted Jovian H20. The H2S is diagnostic of
the cometary sulfur abundance, and possibly of the reservoir of Jovian

SESSION 03 ....... Comet Shoemaker-Levy 9 III
Monday, 1:30 - 3:30 Crystal Ballroom A
Cindy C. Cunningham and Glenn Orton, Moderators

03.01 Mac Low M.-M.* Zahnle K.
Comparison Between Observations and Numerical Models of the
Impact of Comet Shoemaker-Levy 9 on Jupiter

We present our latest numerical models of the impact of comet
Shoemaker-Levy 9 on Jupiter and compare to the observations. We use the
astrophysical gas dynamics code ZEUS to model the initial entry,
explosion, and plume development for a period of 10 minutes after impact
of each fragment. We show that if the incoming fragments had diameters
of order 0.5 km as suggested by Asphaug & Benz, we can give a consistent
model of the observations. Our entry models show that these small
fragments explode in the ammonia cloud layer, explaining the
observations of ammonia but no water. Published models predicting
deeper penetration were flawed by inadequate resolution. The light
curves from the resulting 10^27 erg explosions are as dim as observed by
Galileo. Finally, our plume models show that these explosions are still
large enough to spread cometary material over distances of 10,000 km in
the upper stratosphere, as observed.

This work was supported by the NSF through grant AST 93-22509, and by
NASA through grant NAGW-2379. Computations were perfomed at the
Pittsburgh Supercomputing Center.

03.02 Gurwell M. A.* Muhleman D. O. Philips J. A. Grossman A. W.
Millimeter Imaging of the Comet P/Shoemaker-Levy 9 Impacts on

Observations were made during each transit of Jupiter during the impact
events of comet SL9 using the OVRO millimeter interferometer (located
near Big Pine, CA). The data were taken in the 3.0 mm continuum (July
16, 17, 22, 23) and the 3.38 mm continuum (July 18-21). The latter
observations include spectral correlator data centered on the HCN(1-0)
transition. These OVRO observations are part of a larger effort to
measure the microwave response of Jupiter to the impacts, and include
observations taken at the BIMA millimeter interferometer and the Very
Large Array.

The main source of opacity at millimeter wavelengths is gaseous ammonia,
with the weighting function peaking near 0.8 bars. Pre-impact modeling
had suggested that the fragments of SL9 would penetrate to well below
this level. This penetration would perturb the vertical ammonia
distribution at and near the impact site, affecting the millimeter
brightness temperature in spatially localized regions on Jupiter.

We are currently in the midst of the arduous reduction of this large
dataset, but at this point we see no obvious effects of the cometary
impacts in maps of the continuum brightness temperature. Our current
maps are of poor SNR, due to rapid phase fluctuations of the Earth's
atmosphere; however, we will soon be able to correct this substantially
in the dataset. This will allow us to construct much improved maps of
the continuum of Jupiter, increasing our sensitivity to more subtle
effects that the comets may have had. If our current results stand, the
lack of discernable variation in the continuum brightness temperature
has important implications for the depth to which the fragments
penetrated, on the size of the perturbed regions in the atmosphere at
that depth, and on the structure of the impacting bodies themselves.

In addition, using the spectral correlator data we will set at least an
upper limit for HCN in the stratosphere.

03.03 Owen T.* Griffin M. J. Marten A. Matthews H. E. Naylor D. A.
Davis G. R. Han B. Bockelee-Morvan D. Colom P. Crovisier J.
Gautier D. Lellouch E. Strobel D. Orton G. de Pater I.
Sanders D. Atreya S.
JCMT Observations of the Collision of Comet Shoemaker-Levy 9
with Jupiter

Using heterodyne techniques, we observed Jupiter with the JCMT at
frequencies corresponding to lines of HCN, H2CO and CO. We tracked
seven different impact locations during the entire period of the
collision of the comet. A beam-switching procedure was adopted and we
checked that no emission feature was present at the symmetric northern
position. We will report on the detection of HCN emission lines at the
impact sites visible from Earth during our observing run. The first
detection was that of fragment C on July 17 UT only one hour after
impact. On the following days, we succeeded in observing similar narrow
lines of HCN at the impact locations F, G, H, A, P2 and R. At the end
of our observations, impact sites had accumulated to such an extent that
several of them were usually contained within our 14" or 21" beam, at
354 and 267 GHz, respectively.

We will present a time history of the emitting regions from a careful
analysis of the recorded spectra of the HCN lines as well as of the
other lines we observed. Despite the poor weather conditions, our JCMT
observations are the only submillimeter detections of HCN during the
SL9-Jupiter collision.

03.04 Kostiuk T.* Fast K. E. Zipoy D. Bjoraker G. L. Buhl D.
Espenak F. Romani P. N. Livengood T. A. Goldstein J. J.
Very High-Resolution Spectroscopy of the Jovian Stratosphere in
the Wake of the SL9 Impacts

We will present results from very high spectral resolution (lambda/delta
lambda approximately equal to 10^6) studies of molecular emission from
Jupiter's stratosphere at the latitude of the SL9 fragment impacts.
Observations were conducted just following the impact period (23 July
UT) and several days after the impacts (26 and 29 July UT), to study the
chemical and thermal alteration of Jupiter's stratosphere and the
reestablishment of equilibrium after the impacts. Measurements were made
with the Goddard Infrared Heterodyne Spectrometer (IRHS) at the Coude
focus of the NASA IRTF. Selected spectral intervals were investigated
near 11 micrometers, which include individual rovibrational lines of H20
and NH3, two non-equilibrium species that may be injected into the
stratosphere, as well as emission lines of the equilibrium-state species
C2H6. No significant H20 lines were observed. Ammonia emission lines
were identified and abundances determined from the precise transition
frequencies and true lineshapes measured at this resolution. From the
linewidth it was determined that NH3 was present at pressures less than
10 mbar in the stratosphere near the Q and K impact sites at least 7
days after the last impact. Emission lines of C2H6 were also measured
and showed no significant enhancement, implying no apparent simple
temperature increase in the stratosphere after a few days.
Interpretation of these results regarding NH3 abundance distribution,
lifetime, and Jovian stratospheric temperature will be discussed.

03.05 Orton G.* Friedson J. Yanamandra-Fisher P. Baines K.
Hoffmann W. Dayal A. Deutsch L. Fazio G. Hora J. Hammel H.
Harrington J.
Jovian Atmospheric Structure Investigation in the NASA IRTF SL9

We explored changes in Jupiter's atmospheric structure due to comet
gragments by imaging the planet between 5 and 21 micrometers using the
Mid-Infrared Array Camera (MIRAC2) in the NASA Infrared Telescope
Facility SL9 campaign. Assuming uniform mixing of both CH4 and H2 at
the impact sites several minutes after impacts, we derived temperatures
near the 10-mbar region of the stratosphere via CH4 emission at 7.85
microns and between 100 and 400 mbar in the troposphere via the
collision-induced absorption of H2 at 13.00, 17.2, 17.8 and 20.8
microns. We find stratospheric warming by as much as 4 K at sites
sampled within 30 hours of impact. However, the time scale for cooling
is on the order of days - much shorter than expected from radiative
cooling. Similar observations in the troposphere showed warming by as
much as 4 K at 100 mbar and 1.5 K at 400 mbar. The 400-mbar temperature
perturbations also disappear on a time scale of days, but the largest
100-mbar perturbations cool to roughly half their initial amplitude in
18 days. There are also major differences in the amplitude of the
perturbations associated with the various impact sites. For example, the
E impact site had barely any detectable perturbation of 100-mbar
temperatures. We also mapped cloud pertubations at 5.00 and 8.57
microns and mapped NH3 gas variability near 10.57 microns. We found
little evidence for perturbations of the 600-mbar NH3 cloud or deeper
clouds. On the other hand, most impact sites had strong 10.74-micron
emission arising from NH3 gas transported into the stratosphere, forming
the most persistent thermal infrared signature at the impact sites.

03.06 Griffith C. A.* Orton G. Noll K. Kelly D. Lacy J.
Zahnle K. Bezard B.
Mid-IR Spectroscopy and Ammonia Images of K Impact Site

We have observed the SL9 comet's crash into Jupiter at the NASA Infrared
Telescope Facility (Mauna Kea, Hawaii), with the U. Texas mid-infrared
echelle spectrometer IRSHELL. This spectrometer was used at a spatial
resolution of 1 arcsec and a spectral resolution of approximately 15000.
We detected ammonia emission at wavelengths of 908 and 945 cm^-l over
the K impact site at several times after impact; 10 hours, 5 days and 10
days later. The emission observed within a week of impact can be most
easily simulated by an NH3 mixing ratio of 4.3 x 10^-8, which is
constant in altitude above the tropopause and equal to the value at the
tropopause. For the quiescent atmosphere, the NH3 abundance falls off
rapidly with altitude in the stratosphere as a result of photochemical
destruction. This situation is consistent with a scenario in which the
impactor penetrated at least as deep as the tropopause. The ammonia rose
following the hot gas of the explosion into the the stratosphere, where
we have detected its presence. The emission decreases over a week
timescale, consistent with depletion of NH3 from photolysis and
diffusion. Emission from C2H4 was also observed, consistent with a
mixing ratio of about 5 x 10^-9. We searched for evidence of H2S and HCN
several hours after the K impact and found no signature of these

03.07 Brooke T. Y.* Orton G. S. Crisp D. Friedson A. J.
Bjoraker G.
Near-Infrared Spectroscopy of the Shoemaker-Levy 9 Impact Sites
with UKIRT: CO, NH3, and Haze Layers

Spectra of impact sites in select regions in the 3-5 micrometer 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. The spectral
resolution was 1.5 x 10^4 with the echelle grating, and 10^3 in low
resolution. The slit was approximately aligned with the impact

CO: Echelle spectra of several impact sites were obtained in absorption
lines of CO near 4.7 micrometers. The effects on the CO line depths are
still being assessed but were not large. However, CO emission was
detected at the L site on Jul 20 at 2:20 UT, over 4 hours after impact.

NH(sub)3: Echelle spectra of the K/W impact sites were obtained in a
little studied region: the NH(sub)3 nu(sub)1 Q-branch at 3.0 micrometers
on Jul 28 at 5:00 UT. The sites were brighter in reflected sunlight
than adjacent regions. Different absorption depth changes in the
NH(sub)3 Q-branch compared to lines of other molecules are seen. The
spectra will provide unique new estimates of the effective height of the
haze layer and the NH(sub)3 abundance.

Haze layers: Low resolution spectra of the E, H, and G impact sites at
3.3 micrometers were obtained on Jul 19 at 8:10 UT. Although bright
against Jupiter's CH4 absorption, the actual reflectivities are low,
approximately a few percent.

03.08 Nicholson P. D.* Gierasch P. Goodman J. Hayward T. McGhee C.
Moersch J. Squyres S. Van Cleve J. Matthews K.
Neugebauer G. Weinberger A. Bjoraker G. Conrath B. Orton G.
Principal Results from Palomar Observations of the
Shoemaker-Levy 9 Impacts

Observations of the impacts of the SL-9 fragments on Jupiter were
carried out with the 5-meter Hale telescope at Palomar Observatory using
two instruments: the near-IR Cassegrain InSb camera (lambda 2 - 4
microns) and SpectroCam-10, a mid-IR imaging spectrometer (lambda 5 - 18
microns). Dual-wavelength lightcurves at 3.5 microns and approximately
4.6 microns were obtained for the R impact on 21 July under good
conditions, interspersed with periodic lambda 8 - 14 micron spectra. A
faint signal of approximately 30 sec duration was detected at 2.3
microns (but not at 4.6 microns) at the expected time of the V impact.
No detectable signal was seen from the B or U impact. The near-IR
camera was also used to monitor the reflectivity spectra and temporal
evolution of the stratospheric aerosol clouds resulting from the
impacts. Spectra at 2.0 - 2.35 microns clearly show the signature of
methane but not molecular hydrogen, and suggest an effective aerosol
pressure level of approximately 10 mb. Follow-up observations on 16-19
August show that the brighter impact sites remained distinct, but had
drifted 12-15 degrees eastward relative to System III, suggesting a mean
wind speed of 5 ms^-1 at this altitude and -44 degrees planetocentric
latitude. SpectroCam-10 was used in three modes: narrowband imaging at
7.9, 8.8 and 10.3 microns to monitor the spatial extent and evolution of
the impact-heated regions in the upper troposphere and stratosphere;
low-resolution (lambda/delta lambda = 100) long-slit spectroscopy to map
the perturbed regions at 8-14 microns; and high-resolution (lambda/delta
lambda = 2000) spectroscopy at selected wavelengths for thermal
retrievals and to probe the vertical distribution of CH4, NH3, and PH3.
We will present an overview of the principal results from these
observations. Detailed results and catalogs of the data are presented
in companion papers by Gierasch et al., Hayward et al., and McGhee et
al. This research was supported by a grant from the NASA Planetary
Astronomy Program.

03.09 Gierasch P.* Goodman J. Hayward T. McGhee C. Moersch J.
Nicholson P. Squyres S. Van Cleve J. Matthews K.
Neugebauer G. Bjoraker G. Conrath B. Orton G.
A Physical Interpretation of the SL-9 Impacts Observed from

Post-impact 8 to 14 micron spectra and images of SL-9 impact sites and
the neighboring regions, obtained with SpectroCam-10 on the Hale
telescope, are used to infer atmospheric properties. Temperature in the
middle stratosphere and in the middle troposphere is deduced from
spectral bands near 8 and 12 microns sensitive to methane and hydrogen
opacity. The vertical distribution of ammonia and phosphine is deduced
from spectral regions near 10 microns. Aerosol opacity at 8 microns and
longer wavelengths does not noticeably differ between impact sites and
neighboring regions. A few days after impacts, over areas on the order
of 2000 km in diameter centered on impact sites, the stratospheric
temperatures are elevated, tropospheric ammonia is depleted, and ammonia
and phosphine appear in the stratosphere at concentrations more than two
orders of magnitude above their usual values. From near infrared imaging
data, obtained with the Cal Tech Near Infrared Cassegrain Camera on the
same telescope, we find that aerosols visible at wavelengths near 2.3
microns are at a pressure level of approximately 10 mb.

03.10 Ingersoll A. P.* Kanamori H.
Waves from the Shoemaker-Levy 9 Impacts on Jupiter

Press release images (1,2,3,4) from the Hubble Space Telescope (HST)
reveal narrow circular rings around the G impact site. The rings'
circular shape and outward expansion implies that they are waves.
Contrast is produced by stratospheric particles that appear bright in a
methane filter. Here, following Kanamori (5) and Ingersoll et al. (6),
we compute the properties of waves generated by explosions at various
depths in Jupiter' s atmosphere. The observable properties are group
velocity, pulse width, and amplitude in the stratosphere. Explosions
near the tropopause mainly excite acoustic waves trapped in the sound
channel at the temperature minimum. Deeper explosions mainly excite
gravity waves trapped in the stable layer near the water cloud (6). Both
waves have narrow pulse width, but the acoustic wave travels several
times faster than the gravity wave. Untrapped waves (7) have lower
amplitude and broader pulse width. The most prominent ring in the HST
images matches the speed of the trappped acoustic wave, while the less
prominent ring matches the speed of the trapped gravity wave. The
relative amplitude of the two waves suggests that the impacts did not
reach the Jovian water cloud.

1. Chapman, C.R. Nature, 370, 245 (1994).
2. Cowen, R. Science News, 146, 68 (1994).
3. Kerr, R.A. Science, 265, 601 (1994).
4. Wakefield, J. EOS, 75, 337 (1994).
5. Kanamori, H. Geophys. Res. Lett., 20, 2921 (1993).
6. Ingersoll, A.P. et al. Geophys. Res. Lett., 21, 1083 (1994).
7. Harrington, J. et al. Nature, 368, 525 (1994).

03.11 Mosser B.* Sauvage M. Lagage P. O. Lognonne Ph. Gautier D.
Drossart P. Livengood T. Kaufl H. U. Billebaud F. Marley M.
Belmonte J. A. Roca-Cortes T.
Seismological Consequences of the SL 9 Comet Impact

Thermal images of the Jovian disk have been recorded with the TIMMI
infrared camera based at the 3.6-m telescope at ESO (Chile), in order to
monitor the thermal fluctuations due to the seismic waves excited by the
comet impact (Lognonne et al. 1994). Data analysis of impacts A, B, F
and H is under progress (B has not been detected, and F only
marginally). During the 2 hours following each impact, a quasi-
continuous series of images has been recorded at the rate of about 1
image each 3 seconds. A broad band filter [9-->10.4 micrometer] has been
used, sensitive to the tropospheric levels around the 0.5 bar level.
About 10,000 images of the full disk of Jupiter have been recorded
without nodding, in order to obtain the high acquisition frequency
required for the observation of the high frequency primary waves excited
by the impact. The detection of the arrival times of these waves will be
possible for all impacts greater than 10^20 degrees J (Lognonne et al.
1994). The resulting hodograms will permit to infer the sound speed
profile in the planetary fluid envelope.

Twenty hours observations have been recorded in the 6 days following the
impact period, in order to search for pressure modes (resonant waves
with periods greater than 5.5 minutes). The nodding technique has been
used, with an acquisition rate of about 1 image per minute. The data
will be mixed with the ones obtained in slmilar conditions at the NOT
with the CAMIRAS camera developed by the Service d'Astrophysique
(Saclay) and at the CFHT with the C10 micrometer camera developed in the
Observatoire de Lyon. The combination of the 3 different sets of data
gives 40 hours observations, and is necessary to reduce the window
effect. Data reduction will be similar to the one commonly used in
helioseismology. The seismological observations may provide the first
measurement of the density profile through the whole planetary interior,
and could permit to discriminate between the poorly constrained current
interior models of the planet.

03.12 Lagage P. O.* Galdemard Ph. Jouan R. Masse P. Pantin E.
Sauvage M. Olofsson G. Huldtgren M. Belmonte J. A. Regulo C.
Roca Cortes T. Rodriguez Espinosa J. M. Selby M. Vidal L.
Mosser B. Gautier D.
10 Micrometer Observations of SL9 Impacts with CAMIRAS at NOT

From July 15th to July 27th, Jupiter was imaged with CAMIRAS, the Saclay
mid-IR camera, mounted on the 2.5m Nordic Optical Telescope (La Palma
island). The 10-13 micron filter and the 0.9" PFoV (total field:
57"x57") were used. The weather conditions were fine. Out of the 8
impacts observed (A, E, F, H, L, Q, T, U), five were clearly detected
(A, E, H, L, Q).

Light curves with a time resolution of 1 second, will be presented. Mid-
IR observations are an indispensable complement to shorter wavelength
measurements, for example to obtain indication about the impact site
temperature. We will insist particularly on the brightest impact, the L
impact, which shows first a faint precursor peak (the plume?) around
22:18 UT, followed by a huge spot, at its maximum 12 minutes later. At
this stage, with an intensity of 15 000 Jy, the spot was one of the
brightest mid-IR source in the sky. The brightness decreases rapidly
with, around 22:36, an intringuing plateau lasting for a few minutes.

The prime aim of the observations was the search for thermal
fluctuations associated with seismic waves excited by the comet
fragments. Direct information about the internal structure of Jupiter
can be derived from the arrival times of the waves at various distances
from the impact. The data analysis is underway.

The post impact observations will be used in coordination with the mid-
IR observations at ESO and CFHT to search for Jupiter pressure modes
(see Mosser et al.).

03.13-P Steffes P. G. DeBoer D. R. Smith W. W.
Observations of the Jovian Microwave (5cm) Emission During and
Subsequent to the Collision with Comet Shoemaker-Levy 9

A student-faculty observational project has been undertaken whereby the
5 cm continuum emission from Jupiter is being monitored both before and
after the collision of Comet Shoemaker - Levy 9. Disk-averaged flux
density variations of 0.5% are detectable with the system developed,
which uses a 30-meter radio telescope at the Georgia Tech Woodbury
Research Facility, 65 miles south of Atlanta. This facility has been
developed by a team including over 50 students.

The 5 cm flux from Jupiter is largely thermal (75%) and will be subject
to variations depending on the depth of the comet's penetration, as well
as the redistribution and spreading of microwave absorbers carried up
from the deeper atmosphere by the vertical shock waves. The non-
thermal, synchrotron emission at 5 cm (approximately 25%), could
likewise be affected by the injection of either cometary material or
Jovian atmospheric constituents into the magnetosphere.

As with any continuum measurement, large bandwidths (>500 MHz) are
desirable so as to increase sensitivity. However, spectral crowding in
the centimeter wavelength range has made it more difficult to obtain
interference-free observing spectrum, especially given the large number
of spaceborne transmitters. New approaches for addressing this problem
are presented.

This work was supported by the NASA Planetary Atmospheres Program under
Grant NAGW-533.

03.14-P Grossman A. W. White S. M. Muhleman D. O. Gurwell M. A.
Microwave Imaging of Jupiter's Troposphere During Impact with
Comet P/Shoemaker-Levy 9

We report on the results from a world-wide campaign to acquire high-
resolution, microwave images of thermal emission from Jupiter's
troposphere, before, during, and after impact with fragments of comet
P/Shoemaker-Levy 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 1-5 bars, below the optically thick cloud deck.

Preliminary images at centimeter wavelength clearly show the expected
zone-belt 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 NH3 mixing ratio decrease by
a factor of two from the EZ to the NEB.

At the latitude of the impacts there are no discernable features at a
level of 3-5K within a 3000-5000 km beam-width. This constrains the
change in the average NH3 abundance within this region to value of less
than 10%, although larger changes in NH3 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.

03.15-P Sada P. V. McCabe G. Deming D. Bjoraker G. Jennings D.
Loewenstein R. F.
Continuing Thermal-Infrared Spectroscopic Observations of
Jupiter and Comet Shoemaker-Levy 9 Fragment Collision Sites

The collisions between comet Shoemaker-Levy 9 and Jupiter created large atmosphe

These observations will be used to retrieve temperature and species abundance in

The preliminary results of this study will be presented.

03.16-P Matthews C. N.
Hydrogen Cyanide Polymers from the Impact of Comet
Shoemaker-Levy 9 on Jupiter

Hydrogen cyanide polymers--heterogeneous solids ranging in color from
yellow to orange to red to black--may be among the organic
macromolecules most readily formed within the solar system. The non-
volatile black crust of comet Halley, for example, might consist largely
of such polymers arising from HCN formed by photolysis of methane and
ammonia. It seems likely, too, that HCN polymers are a major constituent
of the dark solids detected spectroscopically by Cruikshank et al. on
other comets and on some asteroids and planetesimals. Comet outbursts
attributed by Rettig et al. to exothermic internal polymerization of HCN
may even be a contributory cause of the fragmentation observed for comet
Shoemaker-Levy 9. HCN polymerization could account also for the yellow-
orange-red coloration of Jupiter.

Laboratory studies by Matthews et al. of these ubiquitous compounds
point to the presence of polyamidine structures synthesized directly
from hydrogen cyanide. These would be converted by water to polypeptides
which can be further hydrolyzed to alpha-amino acids. Other polymers and
multimers with ladder structures derived from HCN would also be present.

The impact of comet Shoemaker-Levy 9 with Jupiter--a collision between
two possible sources of HCN polymers--would be expected to disperse some
polymeric cyanide material as well as to produce pyrolysis and
hydrolysis compounds such as nitrogen heterocycles, nitriles,
polypeptides and amino acids. Spectroscopic detection of these predicted
products and of the parent polymers of HCN would strengthen
significantly the hypothesis that cyanide polymerization is a preferred
pathway for prebiotic and extraterrestrial chemistry.

03.17-P Hayward T. L. Gierasch P. Goodman J. McGhee C. Moersch J.
Nicholson P. Squyres S. Van Cleve J. Matthews K.
Neugebauer G. Weinberger A. Miles J. Orton G.
Ten Micron Images and Spectra of the Shoemaker-Levy 9 Impacts
Observed from Palomar

We present lambda = 5 to 13 micron observations of several SL-9 impacts
made with SpectroCam-10 at Palomar Observatory from July 15 to 24 and
August 15 to 18. On the 5-m Hale Telescope, the diffraction limit of
0.5" at 10 microns provided excellent views of the impact sites. Our
8.8 and 10.3 mmicron broad-band images illustrate the morphological
structure and evolution of several sites from soon after the impact to
about one month later. In many spots the thermal infrared structure
closely resembles the near-IR appearance. Low-resolution 8 to 13 micron
spectra show strong thermal continuum emission in young spots which
slowly decays over the following days and weeks. High resolution
spectra (R = 2000) in lines of H2S, NH3, and PH3 show features such as
line inversions in several of the larger impact sites. We will also
discuss 5 micron images and 8-13 micron low-resolution spectra of the R
impact. The emphasis of this paper will be to catalog and describe the
observational results; detailed physical interpretations will be
presented in companion papers by Nicholson et al. and Gierasch et al.,
and the Palomar near-IR observations will be discussed by McGhee et al.
This research was supported by a grant from the NASA Planetary Astronomy

03.18-P Meadows V. Crisp D. Orton G. Brooke T. Spencer J.
AAT Observations of Shoemaker Levy-9 Collisions with Jupiter

We will use the Infrared Imaging Spectrometer (IRIS) on the 3.9m Anglo-
Australian Telescope (AAT) to observe the collisions of the fragmented
comet Shoemaker Levy-9 (SL9) with Jupiter on July 16-22. This southem
hemisphere site (31.28 S) is advantageous because Jupiter will be at -12
degrees declination, and at least 6 impacts can be observed from this
longitude (149.07 degrees E). The impacts of fragments D, K, N, and W
will occur after sunset, while the impacts of C and G will be visible
before sunset. IRIS is versatile, near-infrared (0.9 to 2.5 micrometer)
camera/spectrometer with a 128 by 128 HgCdTe (NICMOS 2) detector. For
imaging at f/15, the IRIS pixel scales are 0.61 and 1.94 arcsec. K-band
images of comet fragments will be used for astrometry, and to search for
evidence of further tidal disruption before impact. Long-exposure images
will be used to monitor interactions between the SL9 dust veil and the
Jovian magnetosphere and ring. Impact events will be monitored by rapid-
sampling K-grism spectroscopy (2 to 2.4 micrometers). In this mode, IRIS
provides a spatial resolution of 0.6 arcsec/pixel along a 60 arcsec
slit, a spectral resolution of approximately 300, and readout times as
short as 0.5 seconds. For all fragments except K, the slit will be
placed across the limb at the impact latitude to provide time-resolved
spectra of the fireball as it rises above the limb. For fragment K, the
slit will be placed across the disk of Europa, which will be in eclipse.
Once the impact sites rotate into view, drift-scanning will be used to
produce spectral images of the Jovian disk. Images extracted within
strong CH4 and H2 bands will constrain the abundance, vertical
distribution, and horizontal motions of aerosols produced by the
impacts. These image cubes will also be used to search for trace gases
of cometary and tropospheric origin (H20, CO, H2S, etc.) deposited in
the stratosphere by the impacts. Finally, the K-band cubes will monitor
variations in the weak H2 and H3+ auroral emission associated with the

03.19-P Bockelee-Morvan D. Colom P. Despois D. Gautier D. Biver N.
Crovisier J. Encrenaz T. Gerard E. Lellouch E. Marten A.
Owen T. Strobel D.
Molecular Observations of the Jupiter/Comet Shoemaker-Levy 9
Collision at the Swedish/ESO Submillimetre Telescope and the
Nancay Radio Telescope

Millimeter line observations of the Jupiter/comet Shoemaker-Levy 9
collision were performed with the Swedish/ESO Telescope (SEST, La Silla,
Chile) from July 18 to July 23 1994. Using alternatively the 3 mm, 1.3
mm and 0.8 mm receivers, we searched for line signatures of several
species at various impact sites, CO (230 GHz and 345 GHz), HCN (266
GHz), HCO^+ (89 GHz), SiO (86 GHz), SO2 (246 GHz), ArH^+ (246 GHz) and
ArH(sub)3^+ (246 GHz). In addition, in late July, after the impacts, and
in mid-August, we observed the 1667 and 1665 MHz transitions of the OH
radical at the Nancay telescope. A quicklook analysis does not reveal
any obvious detection. Further reductions are in progress. A report will
be presented.

03.20-P Caldwell J. J. Barnet C. Noll K. S. McGrath M. Weaver H. A.
Atreya S. Trafton L. Yelle R.
HST/GHRS Spectra of Jupiter's from 1610 to 1830 Angstrom During
the SL-9 Encounter

The G140L grating in the GHRS and its associated detector (D1) onboard
the Hubble Space Telescope have 500 spectral diodes or width 0.57
angstrom. During and after the SL-9 encounter period, a total of ten
spectra of Jupiter were obtained with the G140L in two spectral ranges:
1251-1537 angstrom and 1541-1828 angstrom. We report here on the four
spectra taken at the latter spectral range, but restrict our discussion
to wavelengths longer than 1610 angstrom. This wavelength approximately
separates shorter wavelengths where emission occurs (predominantly from
H2) from longer ones where reflected sunlight dominates.

Consistent with other UV observations during the SL-9 event, we find
that the general reflectivity of Jupiter is significantly lower than it
is at normal times, Within our restricted spectral range, C2H2 has
previously been determined to be the strongest molecular absorber, and
several characteristic bands have been detected during SL-9 around 1750
angstrom. However other absorbers, possibly continuum and/or molecular,
appear to be present. A continuing investigation seeks to determine the
nature and identity of these additional absorbers.

03.21-P Kundu A. Grossman A. W. de Pater I.
Millimeter Wave Observations of the Effects of Shoemaker-Levy 9
on the Jovian Troposphere

We present radio interferometric images of Jupiter in the 3mm thermal
continuum before, during and after the Shoemaker-Levy 9 collision. The
data were obtained using the Berkeley-Illinois-Maryland array (BIMA), a
six-antenna millimeter interferometer that operates in the 3mm band with
an 800 Mhz wide continuum.

We observed Jupiter continuously from 15-24 July at 101 Ghz with a
bandwidth of 725 Mhz . However we tuned to the HCN line at 88.6318 Ghz
from the 19 to the 21 of July when preliminary reports from other
observers suggested a detection of HCN.

Our synthesized images of Jupiter have an angular resolution of 4
arcsecs, and sensitivity of 1 K. As of yet we have not discovered any
localized or large scale change in the thermal emission in the
millimeter range. We have not found any evidence of a HCN line in onr

Our observations are very sensitive to any vertical motion in the Jovian
troposphere associated with the impact at a level of 0.5-2.5 bars. Thus
our observations can place unique upper bounds to the depth to which
collision induced effects occured. Consequently, useful constraints can
be obtained on the size and composition of the cometary fragments.

03.22-P Vervack R. J. Jr. Collins J. Holberg J. B. Sandel B. R.
Herbert F. Dessler A. J. Forrester W. T. Broadfoot A. L.
Voyager 2 UVS Observations of Jupiter During the Comet
Shoemaker-Levy 9 Impact Events

The Voyager 2 Ultraviolet Spectrometer (UVS) observed Jupiter during the
Comet Shoemaker-Levy 9 impact events. Observations began on July 8 and
continued through August 15 with approximately 18 hours of real-time
data collected each day. Data were collected for impacts A, B, C, D, F,
G, H, L, Q1, Q2, R, T, U, V, and W, as well as during the predicted
impact times for the "missing" fragments J and M. Data from impacts E,
K, N, P1, P2, and S were not collected because they occurred during gaps
in scheduled ground-station coverage. The Voyager 2 UVS is sensitive
over the 500-1700 angstrom range with a spectral resolution of
approximately 18 angstrom. The time resolution of these observations is
3.84 seconds. Although Voyager 2 had a relatively direct view of the
Jovian hemisphere where the impacts occurred, the spacecraft was 40.9 AU
from Jupiter.

An examination of the UVS spectra revealed no obvious signal from the
comet fragment impacts. At present, we can quote a provisional upper
limit of 2x10^35 photons/sec or 2X10^24 ergs/sec integrated over the
wavelength range 1340-1670 angstrom for the radiated power escaping the
Jovian atmosphere. This wavelength range was chosen because it is
relatively free of absorption by atmospheric H2 and the primary
hydrocarbon CH4. We expect to improve this upper limit and to relate
this power to various emission scenarios. This work was supported by
NASA Grant NAGW-3657.

03.23-P Moses J. I. Allen M. Gladstone G. R.
The Aftermath: Effects of Shoemaker-Levy on Jovian

The impact of Comet P/Shoemaker-Levy 9 (SL9) with Jupiter caused
cometary material and tropospheric Jovian air to be injected into
Jupiter's stratosphere. Atomic and molecular species that are not
commonly found in the Jovian stratosphere can now participate in
atmospheric photochemistry. Observations suggest that relatively few
new gaseous oxygen species were introduced to the stratosphere following
the impacts but that sulfur and nitrogen species were greatly enhanced.
We have examined the short- and long-term photochemical evolution of
sulfur and nitrogen compounds in the Jovian atmosphere. We begin our
calculations approximately a half hour after an impact, after the
material in the ejecta plume has splashed back down into the atmosphere.
The results of several shock chemistry models as well as the reports of
the various nitrogen and sulfur species observed immediately after the
impacts provide our initial composition. We find that the sulfur
species observed initially are short-lived -- S2 and CS2 are photolyzed
in less than a day, and H2S in less than a week. Several interesting
sulfur species are formed in the hours and days following the impacts.
Particularly noteworthy are CS, CH2S, CH3SH, CH3SCH3, CH3SSCH3; ring
species such as elemental sulfur, C2H4S, and C3H6S; and polysulfide
chains such as HSxH. The nitrogen photochemistry is somewhat less
intricate. Ammonia (NH3) is photolyzed in about a week, but the
photolysis products are recycled back to NH3 or form hydrazine (N2H4).
Molecular nitrogen is very stable in the Jovian stratosphere and will
not participate in the photochemistry. If HCN is produced in sufficient
quantities during the impact, its photolysis may dominate the production
of nitriles and more complex nitrogen compounds. Many of the sulfur and
nitrogen compounds formed after the impacts will condense in Jupiter's
stratosphere. The atmosphere will take years to relax fully back to its
unperturbed state.

03.24-P Billebaud F. Drossart P. Vauglin I. Merlin P. Sibille F.
Lognonne P. Mosser B. Lagage P.-O. Lellouch E. Gautier D.
10 Micron Observations of Hydrocarbon Emissions in Jupiter's
Atmosphere After the Impacts of Comet P/Shoemaker-Levy 9

Using the French national 10-micron array camera installed on CFHT, we
recorded images in hydrocarbon bands around 10 micron after the impacts
of the fragments of comet P/Shoemaker-Levy 9.

We used the circular variable filter of the instrument which provides a
spectral resolving power lambda/delta lambda similar to 50. Jupiter was
mapped at the following wavelengths: 7.93 microns (CH4, P branch of the
upsilon(sub)4 band), 12.18 microns (C2H6, upsilon(sub)9 band), and 13.7
microns (C2H2 upsilon(sub)5 band). The observations cover the period
from July 24 to July 27 and provided several sequences of observations
at the mentioned wavelengths, allowing us to cover the whole planetary
disk. The spatial sampling was 0.8 arcsec per pixel.

A preliminary analysis of a sequence of observations at 7.93 microns
shows CH4 emissions associated with the impact regions. The spatial
distribution of these emissions, and its temporal evolution will be
characterized. The comparison of the emissions of the different
hydrocarbon bands should help determining the atmospheric levels
affected by the impacts.

03.25-P Knacke R. F. Geballe T. R. Noll K. S. Brooke T. Y.
Infrared Spectra of the R Post-Impact Events of Comet
Shoemaker-Levy 9

We obtained a series of 2.20-2.41 micron spectra of the R event of Comet
Shoemaker-Levy 9 shortly after the impact. Spectra taken with the UKIRT
telescope span a 45 minute interval, spaced roughly 1-2 minutes apart.
Before the event, spectra of Jupiter's limb at 44 deg south showed
reflected solar llght with mainly CH4 absorption. Within about two
minutes the flux rose by approximately two orders of magnitude and the
spectrum developed a classic CO 2-0 bandhead profile at 2.295 microns.
Other CO bandheads were not clearly visible either because they were
weak or were blended with other emission features (probably CH4).
Subsequent spectra, obtained at 1-2 minute intervals, showed a
progressive weakening of the continuum and the CO emmission. The 2-0
profile was absorbed at the long wavelength side, apparently by cooler
CO on the outside of the blast, until finally CO emission and absorption
faded away. The entire episode was over in approximately ten minutes. We
will discuss interpretation and modeling of the spectra. This work was
supported by NASA grant NAGW 2194.

03.26-P Marten A. Moreno R. Paubert G. Wild W. Colom P.
Crovisier J. Rosenqvist J.
New Heterodyne Millimeter Observations of Jupiter Performed
After the Collision of Comet SL9 with the Planet

Emission lines of CO and CS have been observed at 230 and 245 GHz with
the IRAM 30m radiotelescope located near Granada, Spain, during the
comet Shoemaker-Levy 9 collision with Jupiter, (c.f. Lellouch et al.,
this issue).

The IRAM 30m has continued its monitoring of Jupiter at several fragment
locations a long time after the last impacts using the same observing
procedure. We report here on new detections of CO and CS lines,
unambiguously observed _in absorption_ at the same frequencies. We have
checked whether these features might be artefact effects in the
observational procedure and conclude they are not.

We present a summary of all observations carried out after July 28th and
discuss the significant aspects of the recorded spectra. A comparison of
our results with other previous millimeter observations of Jupiter will
be done.

Since the spectral signatures of carbon compounds did not disappear
rapidly, we will propose a coherent scenario of their formation in the
upper atmosphere of Jupiter.

We are indebted to M. Grewing and M. Guelin for providing this unique
opportunity of monitoring Jupiter during extra time periods of
observations with the IRAM 30m Telescope.

03.27-P Formisano V. Bellucci G. Lopez Moreno J. J.
Multispectral Imaging of the Shoemaker-Levy 9 Impacts on
Jupiter from Sierra Nevada Observatory

The impact of comet Shoemaker-Levy 9 on Jupiter has been observed from
the Sierra Nevada Observatory, in Granada, Spain with a 1.5 meter
telescope and two instruments: a CCD camera and a multispectral imaging
system (VNIR) with cooled CCD. Observations have been carried out from
July 16 to July 22. This paper concerns mostly with multispectral images
and related spectra ( 400 - 1050 nm ). The cometary dark spots are
clearly visible; the preliminary study of the spectra shows that the
major effect seen in our data is a decrease of the albedo, or a decrease
of the continuum in the spots compared with the nearby -43 degrees band.
The spectra are essentially coincident around 800 nm, and they diverge
going toward 400 nm. The spot spectra show less deep methane bands and
occasionally deeper ammonia band ( at 640 nm ) with respect to the
nearby regions at the same latitudes. Several other small spectral
differences are being studied.

This work was supported by ASI and CSIC grants.

03.28-P Naylor D. A. Davis G. R. Griffin M. J. Marten A.
Holland W. S. Matthews H. E. Han B. Bockelee-Morvan D.
Colom P. Crovisier J. Gautier D. Lellouch E. Strobel D.
Orton G. Owen T. de Pater I. Sanders D. Atreya S.
Broadband Submillimetre Observations of the Collision of Comet
Shoemaker-Levy 9 with Jupiter from JCMT

In addition to the JCMT heterodyne observations of the collision of
Comet Shoemaker-Levy 9 with Jupiter reported by our group at this
meeting, a polarizing Fourier Transform Spectrometer was used to conduct
a broadband spectral survey of the encounter. Despite the poor weather
experienced throughout the run, spectra were obtained in the atmospheric
windows centred on 750, 850 and 1100 microns. To minimise baseline
variations, which can contaminate the transformed spectra in a Fourier
spectrometer, instead of trying to track the impacts around the planet,
the telescope pointing was fixed at the central meridian of Jupiter at a
latitude determined by the beam size in such a way to minimise baseline
variations due to any tracking errors. In this mode impact sites were
observed as they moved through the telescope field of view.

Bad weather prevented us from obtaining pre-collision reference data at
the impact latitudes. Reference spectra were obtained by observing the
planet at an equivalent offset position in the northern hemisphere.
While the 750 and 850 micron band spectra appear featureless, spectra in
the 1100 micron band reveal a repeatable broad, and therefore possibly
tropospheric, absorption feature of the order of 5% of the continuum
background. These results will be presented and their correlation with
impact sites discussed.

03.29-P Vasilyev V. P. Sergeev V. A.
Calculation of the Comet-Jupiter Collision Wave Response

The Jovian atmosphere wave response belongs to direct consequences of
the comet-Jupiter collision. Relatively slow warm-up with surface
vaporization should prevent to explosion of the fragment as a unit. On
the other hand sufficiently rapid attainment of breaking impact stress
should result in this intensive failure. These above principal
concurrence factors determine the wave disturbances source structure in
the Jovian atmosphere. We assume that the source have a heat nature and
stretched drop-like geometry with longitudinal temperature
inhomogeneity. On the basis of such model using the calculations of
Jovian atmosphere oscillation spectra outside the impact was performed.
The most probable existence of the two main amplitude,increasing regions
in the oscillation power spectra is shown. The first region relates the
wavelengths close to the cutoff frequency and the other one relates to
the periods near the 10 D/C, where D is a fragment diameter and C is the
sound velocity. So far as the cutoff period monotonously increases in
the undertropopause layers the depth of the impact for every fragment
can be determined by the peculiarities of the observed low-frequency

03.30-P Betz A. L. Boreiko R. T. Bester M. Danchi w. C. Hale D. D.
Stratospheric Ammonia in Jupiter as a Result of Comet SL-9

Several spectral lines of ammonia in the 10 micron band were monitored
in Jupiter between July 16 and August 8, 1994, covering the period of
fragment collisions from comet SL-9. Narrow line emission was detected
from impact sites E, G, H, K, L, and Ql, while only upper limits were
obtained at the A and C sites and other selected locations on the
planet's surface outside of the impact areas.

The strongest emission was seen from the G site, where the aQ(2,2),
aQ(6,6), and the aQ(9,8) lines of ^14NH(sub)3 were detected. Over a
period of days following the impact, Rayleigh-Jeans antenna temperatures
were about 2 K for the first two lines, corresponding to a brightness
temperature of 200 K if the emission fills the 1.5 arcsec beam. Observed
linewidths were about 1.7 km/s (FWHM) at the beginning of the period and
closer to 2-3 km/s toward the end. The narrow linewidths are definitive
indicators that the emission occurs well above the tropopause.

The lack of equivalently strong emission from water vapor which is
expected to be much more abundant than ammonia in cometary ice leads us
to conclude that the observed ammonia is an impact induced upwelling of
material from the troposphere of Jupfiter where ammonia has long been
known to be a significant constituent. This result helps establish
minimum penetration depths for impact fragments producing detectable NH3

This work was supported by NASA Grant NAGW-3196 (U. CO) and ONR grant
N00014-89-J-1583 and NSF grants AST-9016474, AST-9119317, AST-9321289

SESSION 04 ....... Solar System Formation I
Monday, 1:30 - 3:30 Crystal Ballroom B
A. P. Boss and W. D. Cochran, Moderators

04.01 Berman A. F.* Kary D. M. Lissauer J. J.
Single Pass Planetary Feeding Zones

We have calculated accretion and rotation rates of planets growing on
eccentric orbits within uniform surface density planetesimal disks for a
variety of orbital inclinations and planet and planetesimal
eccentricities (Berman and Kary 1993, B.A.A.S., 25, 1121). Accretion
occurs from two bands which are nearly symmetrically placed about the
planet's orbit. We have found that the extent of a planet's single pass
feeding zones depends only upon the sum of the eccentricities of the
planet and the planetesimals and their mutual inclination. The inner
band is a bit narrower than the outer band and is located slightly
closer to the planet.

For 2-D simulations, the interior limits move slightly closer to the
planet with increasing eccentricity, and the exterior limits move away
from the planet and become comparable to (e(sub)1 + e(sub)2) a for large
eccentricities. Different simulations with the same maximum relative
eccentricity produce very similar limits of the planet's single passage
feeding zone. Were we to extend our simulations over many synodic
periods, we expect the feeding zone would expand more rapidly for larger
planetary eccentricity (with e(sub)1 + e(sub)2 fixed), as the Jacobi
parameter, which is constant when e(sub)1 = 0, tends to vary more
rapidly as e(sub)1 increases. Inclination reduces the mean effect of the
planet's gravity, and thus moves the interior limits slightly closer to
the planet and the exterior limits significantly closer to the planet.

This research was supported in part by NASA-PG&G.

04.02 Ward W. R.* Hahn J. M. Rettig T. W.
Resonant Trapping in a Self-Gravitating Planetesimal Disk

A planetesimal's orbit will decay due to the solar nebula gas drag. Such
an object approaching a massive secondary (a protoplanet) will
experience strong gravitational perturbations at the secondary's outer
Lindblad resonances. If perturbations can compensate for the orbital
energy and angular momentum lost by the planetesimal due to gas drag,
orbit decay is halted and the planetesimal is trapped in a stable orbit
at the resonance. Past examinations (Weidenschilling and Davis 1985,
Malhotra 1993, Kary et. al. 1993) show an isolated test particle can be
trapped at resonance provided the planetesimal size is larger than a
critical value, or, equivalently, the secondary's mass exceeds a certain

We present results of an alternate treatment of this problem that
considers the collective gravitational effects of numerous disk
particles present at a resonance. As a consequence of the planetesimal
disk's self-attraction, the secondary launches spiral density waves at
its Lindblad resonances. Energy and angular momentum deposited there by
the secondary are transported away by wave action, significantly
reducing its ability to trap particles. Provided the waves are able to
propagate out of the resonance zone and damp to the gas and/or particle
disk before reflecting at at the Q-barrier and returning to resonance,
particle trapping is not possible for a significant range of solar
nebula parameters.

This provision that the waves transport the energy and angular momentum
out of the resonance zone ultimately determines whether particle
trapping can occur. Particle trapping in a self-gravitating disk is
possible only if (i) waves vigorously damp in the resonance zone via
linear or non-linear viscous dissipation, (ii) the particles' dispersion
velocity is too high to support wave propagation, or (iii) gas drag
and/or viscous damping of spiral waves is too weak to prevent their
reflection and return to resonance. In the last instance, the returning
waves are re-absorbed by the particles at resonance.

The range of solar nebula parameters (particularly the secondary's mass
as well as the disk particle size in a unimodal disk, or the particle
size distribution for a more general disk) for which collective effects
prevent resonant trapping will be discussed.

04.03 Dominik C.* Tielens A. G. G. M.
Coagulation in Oblique Collisions Between Micron-sized Grains

It is now well established, that coagulation plays an important role in
the formation of planetesimals in the solar nebula. Especially the first
growth step from (sub)micron sized grains to cm sizes is undoubtly due
to this process. Nevertheless, the physics of coagulation is still
poorly understood. In a recent paper, Chokshi et al. (1993) have studied
the head-on collision between two micron-sized grains and derived
sticking probabilities by considering adhesive forces and elastic waves
excitation. We have now extended this study towards oblique collisions.
In these collisions, tangential forces in the contact area can give rise
to sliding or rolling of the spheres over each other, a process that
proves to be decisive for the structure of the formed aggregates.

We have studied the effects of tangential forces, rolling and sliding
friction in an adhesive contact between two elastic, adhesive spheres.
We find, that both sliding and rolling depend strongly on the properties
of the grain materials. In collisions with initial velocities only
slightly smaller than the rebounce velocity, rolling or sliding can be
sustained until a grain formed several contacts with an aggregate. Thus,
grains will generally not stick where they hit, and the solar nebula
aggregates might have been more compact that previously assumed.

References A. Chokshi, A.G.G.M. Tielens, D. Hollenbach, 1993, APJ 407,

04.04 Weidenschilling S. J.*
Formation of Cometesimals and Planetesimals: Numerical Results

The analytie model of formation of "rubble pile" comets (Nature 368,
721) envisions a two-stage process of eollisional coagulation followed
by gravitational instability of a layer of macroscopic bodies. The
instability mechanism requires that the layer exceeds a critical density
delta*; also, the drag-induced radial velocity dispersion must be
sufficiently small. Numerical simulations including time-dependent
settling and coagulation, particle size distribution, and shear-induced
turbulence are in good agreement with analytical predictions. Starting
from dispersed dust at 30 AU, instability is reached after ~-9 x 10^4 y
when the median particle size is ~-15 delta*. About 2/3 of the total
solids are in the dense layer, and the optical thickness has declined to
~-10% of the initial value.

Application of the same model to silicates at 3 AU results in
instability after ~-10^4y for a layer of 40 m bodies containing ~-1/3 of
the total surface density of solids. If local shear is the only source
of turbulence, it delays but does not prevent gravitational instability.
Global (e.g., convective) turbulence corresponding to alpha <~10^-6 has
little effect. Turbulence with alpha >~ 10^-4 has serious consequences;
thickening of the particle layer lowers the coagulation rate and also
keeps the deviation of the gas from keplerian rotation at a large value,
resulting in rapid orbital decay. If the solar nebula contained regions
of high and low turbulence, there would be a strong tendency for solids
to accumulate in the low-turbulence regions.

04.05 Cassen P.*
Meteoritic Constraints on Models of the Solar Nebula: The
Abundances of Moderately Volatile Elements

The "moderately volatile" elements are those which condense (or
evaporate) in the temperature range 650 - 1350 K, as a mix of material
with solar abundances is cooled (or heated) under equilibrium
conditions. Their relative abundances in chondritic meteorites are solar
(or "cosmic", as defined by the composition of CI meteorites) to within
a factor of several, but vary within that range in a way that correlates
remarkably well with condensation temperature, independent of chemical
affinity. It has been argued that this correlation reflects a
systematically selective process which favored the accretion of
refractory material over volatile material from a cooling nebula. Wasson
and Chou (Meteoritics 9, 69-84, 1974, and Wasson and co-authors in
subsequent papers) suggested that condensation and settling of solids
contemporaneously with the cooling and removal of nebular gas could
produce the observed abundance patterns, but a quantitative model has
been lacking. We show that the abundance patterns of the moderately
volatile elements in chondritic meteorites can be produced, in some
degree of quantitative detail, by models of the solar nebula that are
designed to conform to observations of T Tauri stars and the global
conservation laws. For example, even if the local surface density of the
nebula is not decreasing, condensation and accretion of solids from
radially inflowing gas in a cooling nebula can result in depletions of
volatiles, relative to refractories, like those observed. The details of
the calculated abundance patterns depend on (but are not especially
sensitive to) model parameters, and can exhibit the variations that
distinguish the meteorite classes. Thus it appears that nebula
characteristics such as cooling rates, radial flow velocities, and
particle accumulation rates can be quantitatively constrained by
demanding that they conform to meteoritic data; and the models, in turn,
can produce testable hypotheses regarding the time and location of the
formation of the chondrite parent bodies and the planets.

04.06 Wetherill G. W.*
Formation of Extra-Solar Terrestrial Planets

During the past decade significant progress has been made in
quantitatively modeling the formation of planetary systems with surface
densities and stellar masses similar to those of our solar system. These
models have been successful in explaining a number of observable
features, particularly in the terrestrial planet and the asteroidal

These same models have been extended to include a range of stellar
masses and preplanetary disk densities. Disk surface densities,
lifetimes, and temperature distributions are not well defined
theoretically or observationally, but it is likely that they are quite
variable even for young stars of the same mass. This investigation
represents a tentative study of the sensitivity of planetary systems to
variations in these parameters.

150 simulations have been made, based on the Monte Carlo model described
earlier (Wetherill, Icarus 100, 307, 1992), for stellar masses of 0.5,
1.0, and 1.5 solar masses. These are grouped according to several sets
of calculations:
1. Surface density independent of stellar mass.
2. Surface density proportional to stellar mass.
3. One solar mass stars with varying surface density.
4. Variation of the position of "Jupiter."
5. Variation in the power law dependence of the surface density.

It is found that in general the number of final planets (~4) interior to
5 AU is insensitive to variation in these assumptions. With a "Jupiter"
at 5 AU, Earth-Venus mass terrestrial planets are concentrated between
0.7 and 1.3 AU, with little dependence on stellar mass. Higher surface
densities tend to lead to larger planets, sometimes ~2 Earth masses. A
more distant Jupiter permits formation of large planets in the
asteroidal region, whereas a less distant Jupiter inhibits the formation
of terrestrial planets. One solar mass stars usually have an Earth mass
planet in the habitable zones of Kasting et al (Icarus, 101, 108, 1993).
Both the larger and smaller stars appear to be less likely to have
habitable terrestrial planets, but they are not completely excluded.

This work was supported by NASA grants NAGW-1969 and NAGW-3928.

04.07 Canup R. M.* Esposito L. W.
Formation of the Moon from an Impact-generated Disk

In light of geological and dynamical evidence, the "Giant-Impact"
scenario has emerged as a nearly consensus view of lunar origin.
Detailed simulations of the impact of a Mars-sized body with Earth have
demonstrated the plausibility of formation of a proto-lunar disk
following a large impact event. A series of works by a group composed of
Cameron, Benz, Slattery, and Melosh (1986,1987,1989, 1991,1994) have
modeled the impact event and subsequent ejection of material into orbit
about the Earth. Their work predicts radial disk profiles, which are
relatively insensitive to the specifics of an impact, and finds that
most ejected material is typically located within the classical Roche
limit for silicate densities.

While the formation of an impact-produced disk has been convincingly
modeled, subsequent lunar formation from the disk has not. Indeed, the
processes leading from the creation of a protolunar disk to its
accumulation into a single, large Moon are not obvious, since
protosatellite disks in the outer solar system have evolved into systems
of multiple moons and rings. Past works have proposed that multiple
moonlets which formed at various radial locations in the protolunar disk
evolved inlo crossing orbits via tidal interaction with the Earth,
(e.g., Cameron 1986). Tidal orbital evolution requires that the
innermost "moonlet" which, forms in the disk be the most massive in
order for all moonlets to eventually evolve into colliding orbits (e.g.,
Cameron 1986, Cameron and Benz, 1991). This is contrary to expectations
from Canup and Esposito (1994), who find that tidal effects in a region
extending approximately one planetary radii beyond the classical Roche
limit greatly limit and alter accretional growth processes. In the
"Roche Zone", like-sized bodies cannot remain gravitationally bound as
they physically overflow their mutual Hill sphere. Accretion in the
inner protolunar disk would therefore be expected to result in a
distribution of multiple small moonlets (who are precluded from
accreting with one another due to the tidal forces of the planet), while
accretion in the outer disk would be much more efficient. This implies
that the largest bodies in an impact-generated disk would naturally form
in the outer disk, which is contrary to the condition needed for all
moonlets to eventually evolve into crossing orbits to form a single

We will present results of our tidal accretion model applied to
realistic initial conditions in an impact-generated disk. Conditions
that could lead to the accumulation of the disk into a single body will
be discussed. This work is supported by the Cassini Science Project and
the Patricia Roberts Harris fellowship program.

04.08 Lissauer J. J.*
Analytic Calculations Relevant to the Origin of the Rotation of
Giant Planets

Jupiter and Saturn. obtained most of their rotational angular momenta
via hydrodynamic accretion of gas from the protoplanetary disk. Detailed
calculations of such flows are quite complicated, and cannot at present
yield unique, generally applicable results because of substantial
uncertainties in the properties of both the protoplanetary disk and the
growing giant planets themselves. Fortunately, it is possible to make
some simple analytic estimates of the spin angular momentum accreted by
a gas giant planet. These calculations ignore planetary gravity, which
may be far less important than it is for the rotation of terrestrial
planets because growing gas planets occupy a large fraction of their
Hill Spheres, i.e., they are distended "fluffy" objects. The results are
illuminating because they provide both order of magnitude estimates of
the amount of rotational angular momentum attainable by growing giant
planets and qualitative measures of the sensitivity of giant planet
rotation to pressure and surface density variations in the
protoplanetary disk and the orbital eccentricity of the growing planet

This work was supported by NASA PG&G grant NAGW-l 107.

04.09 Foster P. N.* Boss A. P.
Shock-Induced Collapse of the Presolar Nebula

We model the formation of the early solar nebula through shock-triggered
collapse of a molecular cloud core. We examine the plausibility of
initiating collapse with the shock wave from an asymptotic giant branch
(AGB) star wind. This choice for the shock is motivated by meteoritical
evidence. Noble gas isotopic anomalies in SiC presolar grains point to
synthesis in an AGB star. Furthermore, Ca,Al-rich refractory inclusions
have isotopic anomalies that can be explained by synthesis in an AGB
star no more than several million years before the formation of the
inclusion. This suggests passage of an AGB wind through the presolar
molecular cloud near the time collapse began. Previous studies of clouds
struck by shock waves have generally found that supernovae are not a
viable triggering mechanism, because such energetic shocks shred the
cloud through Rayleigh-Taylor (RT) and Kelvin-Helmholtz (KII)
instabilities (e.g., Klein, McKee, and Colella 1994). Our first goal is
therefore to determine if AGB shocks can induce collapse without
destroying the cloud. Secondly, we follow the shock/cloud interface to
determine if this scenario can inject isotopically anomalous grains from
the AGB wind into the solar nebula. Boss (1994) and Boss and Foster
(1994) used 3D hydrodynamical models to show that AGB winds are capable
of inducing collapse and injecting significant amounts of grains.
However, these 3D models were limited in their ability to follow the
instabilities that tore apart the clouds in the supernovae shock
simulations. Here, we calculate much higher spatial resolution 2D
models, which are capable of following the RT and KH instabilities.
Further, we employ a new hydrodynamics code (VH-1:Lindahl) which was
designed to handle shock waves (Collela and Woodward 1984). With this
code, we have reproduced the supernova-shock cloud shredding results.
Preliminary results for the AG13-shock are promising for initiating
collapse. Shocking a Bonnor-Ebert sphere (n(sub)c = 10^6cm^-3) with a 20
km/s AGB wind results in the cloud being crushed onto the symmetry axis.
By the time the shock has passed the cloud, roughly 0.7 cloud crushing
times, the central density has increased by a factor of 5, high enough
to ensure sustained collapse. This research was partially supported by
NASA grants NAGW-3413 and NAGW-1410.

04.10 Ruzmaikina T. V.*
Reprocessing of the Presolar Dust Aggregates in the Accretional

Laboratory studies of a variety of extraterrestrial materials provide an
evidence that the inner solar system, including most of the asteroid
belt, had been subjected to temperatures 1200 - 1400 K. A large amount
of the organics in the dust of comet Halley, and isotopic anomalies in
the outer solar system, assume that the outer early solar system might
preserve significant amount of the interstellar dust, including ices.
This paper studies impulsive heating of the presolar dust and gas in the
accretional shock produced by infalling gas and dust at the surface of
the solar nebula.

The peculiarity of the shock in the solar nebula is fast cooling of the
gas, resulting in a sharp density gradient in the postshock region, and
the density increase up to two orders of magnitude. Millimeter-size
dense grains, embedded in the infalling gas, have such an inertia that
they cross the region of cooling without deceleration, and are heated by
the drag through the cooled and compressed postshock gas. In the inner
part of the solar nebula, up to distances of the asteroidal belt, these
silicate grains are heated to > 1600 K were melted and then solidified
in ,~ 10^3 s, plausibly forming chondrules.

In the outer part of the solar nebula, where the temperature of the
infalling gas is low enough to preserve presolar organic material and
ices, the accretional shock can cause melting, evaporation, and chemical
changes of volatile components of dust aggregates. Also, the presence of
liquid water or vapor could cause oxidation of less volatile components,
such as Fe. The intensity of heating of aggregates and the extent of
region of their evaporation in the outer solar system are also dependent
on the size and structure of the aggregates. Large dense grains of ice
could be melted at distances up to >= 30 AU. Smaller (say <=
micrometer-size grains) and very low dense dust aggregates are heated to
lower temperatures. Such particles preserve interstellar water ice
beyond 5 to 10 AU, and organics beyond 3 to 5 AU. The outer solar system
must contain a mixture of interstellar ices, and ices which were
evaporated in the accretional shock and recondensed again.

04.11 Dyt C. P.* Prentice A. J. R.
Modelling of Supersonic Turbulent Convection in the Solar

A new method of flux corrected transport in two dimensions has been
developed Wo by extending the work of Zalesak (J. Comp. Phys. 31, 335,
1979). The code has been applied to the problem of turbulent
compressible convection in the solar atmosphere with a view to
understanding the origin of the planetary system. The numerical
simulation is carried out for an ideal gas on a uniform mesh between two
horizontal stress-free boundaries. The initial density-temperature
gradient is chosen to be strongly super-adiabatic, corresponding to
polytropic index m = 1. The initial density ratio between the lower and
upper boundaries is 11:1. This thermal instability quickly leads to a
structure consisting of several convective cells dominated by narrow
fast-moving downflows and broad slower-moving M upwellings. Supersonic
speeds and shocks occur near the head of the downflows. C The flows are
time dependent and persist for several sound crossing times. Al These
calculations confirm the findings of Cattaneo et al. (Astrophys. J. 349,
L63, 1990) that supersonic convection can develop in stellar
atmospheres. The most important discovery we report here is that
supersonic convection induces a E concentration of the gas towards the
lower boundary (z = 1) and a rarefying near the top (z = 0). For the
typical density profiles shown below, corresponding to a Rayleigh number
of 4.76 x 10^5, the average density ratio between the lower and upper
boundaries is 17:1. The peak Mach speed is about 1.5. These results
confilm a basic premise of the theory of planetary system origin due to
one of us (AJRP, Astr. Astrophys. 27, 23.7, 1973; Moon & Planets 19,
341, 1978) that supersonic convection caused the contracting proto-solar
cloud to become very centrally condensed. This work is supported by the


? Al

04.12 Chamberlain J. W.*
Blowoff of Planetary Atmospheres: Supersonic Solutions of the
Solar-Wind Equations

We re-examine the inviscid solar-wind equations with heat conduction
from below, and establish a fundamentally new approach for finding solar
and planetary solutions. Although the problem is fourth order, only two
independent integration constants can be assigned, since two boundary
conditions that are required to specify well-behaved supersonic
solutions determine the values of the other two constants. The Noble
and Scarf (1963) solar-wind models are essentially accurate for
practical purposes, but in a fundamental sense they are not

The ratio of thermal energy to gravitational potential at the supersonic
point, (kTr/GMm)s, must lie between 0.4375 and about 0.405 to permit
stable supersonic solutions for ionized hydrogen. For planets with
diatomic molecular atmospheres, the range is lower but still quite
narrow. These limits seriously constrain the physical conditions
wherein hydrodynamic "blowoff" of planetary atmospheres can develop. In
addition, an atmosphere having blowoff conditions is adiabatically
unstable just above the sonic level.

04.13-P Kary D. M. Lissauer J. J.
Gas Drag and the Origin of the Trojan Asteroids

The origin of Trojan asteroids, which librate about Jupiter's L(sub)4
and L(sub)5 points, presents a long-standing puzzle in planetary
dynamics. Models of Trojan asteroid origin are usually based on the
assumption that they were inserted into their present orbits as a result
of collisions after the formation of Jupiter. However, Peale (1994,
Icarus 106 308) has shown that the Lagrange points could be stable in
the presence of the solar nebula, opening up the possibility that
nebular gas drag played a role in inserting asteroids into the Lagrange
points. We find that a planetesimal can have a close encounter with a
planet which inserts it into the 1/1 resonance. When the planet is on a
circular orbit, the encounter leaves the planetesimal in a low
eccentricity orbit (due to conservation of the Jacobi constant) which
allows a second close encounter with the planet. The symmetry of the two
encounters means the planetesimal is ejected into an inferior orbit.
However, when the planet has an eccentric orbit, the planetesimal can be
inserted into the resonance with a large eccentricity. This in turn
introduces an asymmetry into the planetesimal's orbital evolution and
prevents the second close encounter. The planetesimal instead goes
through a decaying libration about the L(sub)5 point. In the case of
planetesimals approaching a 10^-6M solar mass planet with an
eccentricity e >/= 0.02, then ~ 10% of the migrating bodies can be
thrown into the 1/1 resonance and decay down to the L(sub)5 point.
Further collisional evolution is still required to redistribute material
to the L(sub)4 point.

This work was supported in part through NASA Planetary Geology and
Geophysics Grant # NAGW-2061 at UCSB and # NAGW-1107 at SUNYSB.

04.14-P Sasaki S. Nagahara H. Kitagami K. Nakagawa Y.
Shock Heating During Solar Nebula Formation and Origin of
Isotopic Fractionation in CAI and Chondrules

Mass-dependent isotopic fractionations of Mg, Si, O have been observed
in some CAI grains. Recently mass-dependent Mg isotopic fractionation is
also found in olivine condrules of Allende meteorites by SIMS study [1].
Large Mg isotopic fractionation is considered to have been caused by
evaporation or condensation process but conditions (T, P, t) have not
been discussed in detail.

Shock heating process through the vertical accretion of cloud gas in the
solar nebular formation is pursued numerically solving ratiative
transfer precisely with frequency dependence [2]. From a widespread
initial density distribution, peak temperature could be higher than
1700K at the asteroidal zone (a = 2.8AU). Outward radiation from shocks
heats up the less-dense off-disk region, where pressure is 10^-8-10^-7

Evaporation experiments of forsterite (Mg2SiO4) at various PH2
conditions [3] show that evaporation rate at PH2 > 10^-5 [bar] is much
higher than that at PH2 < 10^-6 [bar]. When the evaporation speed is
comparable to the diffusion speed, enrichment of heavier isotopes in
evaporation residues is expected. At a low pressure (PH2 < 10^-6 [bar]),
longer duration (10^4-10^5 min at 1700-1800 K) (or slower rate of
temperature change) is preferable for the isotopic fractionation.
Typical distance for Mg isotopic fractionation is in the order of 10
micrometers, which would be compatible with size of precursor grains of
CAIs and chondrules-before aggregation. Since the evaporation rate is
likely to be higher than Mg diffusion rate at a higher pressure (PH2 >
10^-5 [bar]), Mg isotopic fractionation would not take place.
Evaporative fractionation of Mg isotopes should have proceeded in
less-dense environment, which is compatible with the shock heating
during solar nebula formation.

References: [1] Koga A. et al. (1994) LPSC XXV, 723-724. [2] Kitagami K.
(1994) Master thesis, Univ. Tokyo. [3] Nagahara H. (1994) LPSC XXV,

04.15-P Levy E. H. Ruzmaikina T. V. Ip W.
Tidal Origin in the Solar Tilt

In a view of the flatness of the solar system, the origin of the
substantial deviation between the solar rotational axis and the angular
momentum vector of the planetary system (approximately 7 degrees) needs
a special explanation. The very inhomogeneous distribution of angular
momentum in the solar system suggests that there was significant
exchange of angular momentum and mass between the Sun and the solar
nebula. Such an exchange tends to keep the rotational axis of the Sun
normal to the plane of the protoplanetary disk.

We consider the possibility that the plane of the protoplanetary disk
became tilted, with to respect to its initial position coinciding with
the equatorial plane of the Sun, as a result of tidal forces exerted by
a passing star(s) early in the history of the solar system.

The figure, which appears here in the hard copy, shows the magnitude of
tilt produced by a one solar mass star moving at 5 km/s and at an angle
of 45 degrees with respect to the undisturbed disk. The vertical axis
gives the final ratio of the Y and Z components of angular momentum as a
function of the impact parameter R (sub) O(shown on the horizontal
axis); the radius of the disk is taken to be 50 AU.

The initially thin disk of planetesimals is thickened as the orbital
inclinations are pumped by the encounter with the star; the resulting
disk thickness increases outward. Because of the consequent increase of
encounter velocities in planetesimal collisions, such an encounter could
have a significant influence on subsequent planet accumulation in the
outer solar system, possibly halting accumulation of a planet beyond the
orbit of Neptune. The probability of such an encounter could be high if
the solar system formed in a dense molecular core with a high efficiency
of star formation.

This work was supported in part by grants from NASA.

SESSION 05 ....... Comet Shoemaker-Levy 9 IV
Monday, 4:00 - 5:30 Crystal Ballroom A
Kevin H. Baines and Philip Nicholson, Moderators

05.01 West R. A.* Friedson A. J. Baines K. H. Seymour M.
Karkoschka E. Hammel H. B.
Wide-Field/Planetary Camera 2 Observations of Jupiter's
Stratospheric Haze and Ammonia Cloud: Post Impact

We observed Jupiter in July and August of 1994 with the Wide
Field/Planetary Camera 2 on the Hubble Space Telescope. The first
observations were obtained just prior to the first of the large impacts
by fragments of comet Shoemaker-Levy 9. The final images occurred a
month after the last of the large impacts. We obtained images at UV
wavelengths (218, 255, and 336 nm), blue (413 nm), near-IR continuum
(953 nm), and in the 889-nm methane band. The principal scientific
objectives of this program are to use the cloud particles formed by the
impacts as tracers of the Jovian stratospheric circulation and as a tool
for studying particle microphysical processes in the stratosphere. The
design of our observations emphasizes emission angle coverage (repeated
observations as Jupiter rotates) and coverage at wavelengths (UV and
methane bands) sensitive to vertical location, over time scales from
hours to weeks. Our first attempts to model some of the dense core
regions indicate that the debris material, which is dark at all
continuum wavelengths, is consistent with particles in the size range
0.15-0.3 micrometers mean radius having an imaginary refractive index
that varies from about 0.006 in the near-IR to about 0.02 at 270 nm. In
the dense core regions the optical depth of the particles is in the
range 5-2 depending on whether the material is distributed between a few
mbar and 400 mbar, or only goes as deep as 200 mbar. The material in
the dense core regions is distributed at least as deep as 200 mbar and
at least as high as a few mbar. There is a decrease in optical depth by
about a factor of 2 within a few days after impact, but substantial
optical depth remains in the stratosphere in late August.

05.02 Baines K. H.* Friedson A. J. Orton G. S. Yanamandra-Fisher P.
Drossart P. Esterle P. Hammel H. Harrington J. Lisse C.
Miller S. Achilleos N. Momary T. Noll K. Shure M.
The Effect of SL9 on Jupiter's Vertical AerosolStructure:
Observational Results from IRTF Near-Infrared Imaging

Near-infrared images of Jupiter have been analyzed to determine changes
of the planet's vertical aerosol structure wrought by the SL9 impacts.
The use of multiple wavelengths in regions of varied atmospheric
molecular absorption between 1.6 and 3.4 microns allows us to derive
aerosol opacities and altitude distributions from the high stratosphere
to the 3-bar level of the troposphere. The impact sites are detectable
in all absorption bands with reflectivities many times higher than from
clouds in the pre-impact atmosphere. This includes the 3.3-micron CH4
fundamental band where no aerosols have been previously detected,
indicating that aerosols are present at the levels of several mbar total
pressure. Our preliminary analysis concentrated on analyzing images in
the 2.14-micrometer H2 collision-induced fundmental band taken over
several weeks. From July 21 to August 7, the spatial extent of the
well-isolated Q1 feature changed by less than 10% (aproximately 900 km),
comparable to the fluctuations in night-to-night seeing. During this
same time, the brightness of the cores of G, L, and Q1 did not change
more than 10%, but the brightness of the H impact site decreased by 20%.
A vertical aerosol model structure with bright 0.25-micron particles
distributed uniformly between 1 and 200 mbar is consistent with both our
near-infrared observations and with visible/red Hubble Space Telescope
images (West et al., 1994, BAAS this issue). Specifically, in the core
of the G impact site, we find particle column densities of 3.0 +/-0.5 x
10^8 cm-2, corresponding to optical thicknesses of 2.4 +/-0.4 and 0.085
+/-0.015 at 0.89 microns and 2.14 microns, respectively.

05.03 Harris W. M.* Ballester G. E. Barker J. Clarke J. T.
Combi M. Jessup K. L. Kozyra J. Vincent M. Budzien S.
Emerich C. Prange R. Gladstone G. R. Hall D. T.
Fireman G. F. Livengood T. A. McGrath M. A. Talavera A.
Woodney L.
Early Results from the IUE Shoemaker-Levy Observing Campaign:
Temporal Evolution of the FUV/NUV Albedo in the Impact Regions

The collision of the remnants of comet Shoemaker-Levy 9 with Jupiter
appears to have had significant effects on the structure of the Jovian
upper atmosphere that were caused by the passage of fragments through
high altitude regions, and by the upwelling of debris and gas from the
lower atmosphere. We present the initial results of our IUE observing
program to monitor the temporal development of the impact sites in the
first hours after a collision. During this period significant evolution
of the chemical composition of the upper atmosphere is believed to occur
as the upwelled material cools and mixes with the background atmosphere.
Our observational technique made use of the IUE small aperture to obtain
a series of quick spatially isolated spectra of the E, G, K, S, and W
impact sites while they were on the approaching limb, at the central
meridian, and rotating out of the field of view. Spectra were obtained
using both the SWP and LWP cameras, with exposure times optimized to get
good signal to noise in the 1700-1950 angstrom, 2300-2550 angstom, and
2500-3200 angstrom band passes. Because of the 3" diameter of the
aperture and the 5" resolution of the IUE telescope, these data provide
the general characteristics of the impact sites rather than the detailed
structure near the center. We will show results indicating large scale
changes in the albedo during the first 5 hours after an impact. This
places a lower limit on the initial size of the affected region, and
provides information about the nature of the absorbing material. We also
present results regarding the rate of evolution in the spot albedo as a
function of wavelength, and discuss the longer term evolution of
selected impact sites.

This work is funded by Nasa grant NAGW-4009 to the University of

05.04 Stapelfeldt K.* Ballester G. Clarke J. Harris W. Trauger J.
Crisp D. Brooke T. Weaver H. Prange R. Emerich C.
Bertaux J.-L. Ben Jaffel L.
HST FUV Imaging of Jupiter's Upper Atmosphere Around the Time
of the Comet Impact

During the impact of comet Shoemaker-Levy-9 with Jupiter a combination
of gases from Jupiter's lower atmosphere, cometary matter, and their
photochemical by-products were deposited in the upper atmosphere. Due to
the presence of scattering haze, dust, and UV absorbing molecules the
sites of the fragment impacts appear dark in the reflected sunlight over
1400-2100 angstrom in images obtained with the Wide Field Planetary
Camera 2 (WTPC2) using the Na Wood's filter (F160W). Due to Rayleigh
scattering by molecular hydrogen, far UV images probe only the upper
atmospheric levels (down to around the 100 mbar levels, above the low
stratosphere haze layer). Motion of the dark clouds provides information
on the magnitude and direction of the circulation in these high altitude
regions. During the impact week, evolution in the morphology and
contrast of the clouds was observed. Material was carried zonally by the
upper atmospheric winds, starting to form a dark band at the impact
latitude, and latitudinal motion was also observed that differed between
impact sites. The dark clouds associated with the impact sites of
fragments A, C and E were observed the most. One interesting feature
appeared about two weeks after the first impact, where some material
associated with impact sites A and E seemed to have migrated north and
to be somewhat detached from the rest of that concentrated at the impact
latitude. This may reflect the effect of different wind patterns/speeds
at the different upper atmosphere altitude levels. In addition to the
comet related phenomena, the WFPC2 far-UV images obtained in May 1994
with our Guaranteed Time Observer program, as well as Guest Observer
images obtained during the comet campaign, have revealed various
features in the solar light reflected by the Jovian upper atmosphere: a)
banded structure that does not exactly follow the visible bands in the
lower atmosphere, b) high altitude dark material associated with the
Great Red Spot, and c) some structure in the dark polar haze.

This work is supported by STScI grant GO-5624,18-93A to U. Michigan.

05.05 Clarke J.* Ballester G. Trauger J. Stapelfeldt K. Crisp D.
WFPC2 Investigation Definition Team Prange R. Emerich C.
Ben Jaffel L. Bertaux J. L. Weaver H.
Far-UV Imaging of Jupiter's Aurora with the HST/WFPC 2 Before
and During the Impacts of Comet Shoemaker/Levy

A series of far-UV images of Jupiter have been obtained with the Wide
Field Planetary Camera 2 on the Hubble Space Telescope using the Na
Wood's filter. The first UV images were obtained on 19 May 1994 over
bandpasses including 1150-2100 angstrom, 1300-2100 angstrom, and 1650-
2100 angstrom. The images show bright aurora at both north and south
poles, with emissions from H Ly alpha (1216 angstrom) and the H2 Werner
and Lyman band series (1150-1650 angstrom), in addition to Rayleigh
scattered solar continuum across the disk with pronounced polar
darkenings which are asymmetric between the north and south polar
regions. There is considerable structure in the north aurora, with a
complete auroral oval and additional emission inside the oval. The
observed emissions are up to approximately 100 kR in brightness, with a
limiting sensitivity of roughly 1-10 kR. A more extended series of
shorter exposures was obtained on 31 May 1994 with 10 min. exposures
covering roughly 5 hours (1/2 Jovian rotation). These images show
localized auroral emission structures which change dramatically on short
time scales, with features which are mainly co-rotating with Jupiter.
Additional far-UV images of Jupiter were obtained over 7 days during the
impacts of Comet Shoemaker/Levy. These images show pronounced darkenings
in the reflected solar continuum at the impact sites, with spatial
structure which evolves with the local upper atmospheric winds over the
week-10 days following the impacts. Jupiter' s aurora are apparent in
each far-UV image, with the northern oval and changing emission
structures along and inside the the oval. In addition, we detect UV
emission from the foot of the Io flux tube on each occasion when Io is
on the sunlit side of Jupiter. During the impact week two image series
revealed auroral emission structures that may be connected with the
impact events: northern auroral arcs near the magnetic conjugate point
from the K impact site roughly 45-54 min. after the impact, and a
variable southern emission feature apparently associated with the R
fragment before its impact. This work is supported by grants STScI GO-
5624.18-93A and JPL 959122 to the Univ. of Michigan.

05.06 Hall D. T.* Gladstone R. Herbert F. Lieu R. Thomas N.
EUVE Observations of Jupiter: Effects of Comet SL-9

The Extreme Ultraviolet Explorer satellite conducted extensive
observations of the Jupiter system before, during and after the arrival
of the fragments of Comet Shoemaker-Levy 9. Approximately 500,000
seconds of data were acquired in 1994. Previous observations of the
Jupiter system revealed that electron-impact generated line emission
from oxygen and sulfur ions resident in the Io plasma torus dominate the
spectrum in the 300-730 angstrom range; these emissions dimmed
moderately after the impact events, and no new torus emissions were
detected. Unexpectedly, following several of the impacts, EUVE detected
transient brightening events in planetary atomic helium resonance
emission, HeI 584 angstrom, which was not detected before the impacts (I
< 1 Rayleigh), although it was measured by the Voyager UVS instruments
during 1979 (I is about equal to 4 Rayleigh). After the impacts of
fragments H and K the HeI 584 angstrom brightness increased to about 25
Rayleigh, (arbitrarily assuming that the atmospheric He enhancement
covered 10% of Jupiter's disk). 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
would have absorbed the emitted HeI 584 angstrom photons.

05.07 Schulz R.* Encrenaz Th. Stuwe J. A. Wiedemann G.
Spatial Distribution of Jovian Near-IR Emissions Associated
with the SL-9 Impact: Evidence for Northern Counterparts

Between July 16 and 31, 1994, Jupiter was monitored in the nearIR with
the IRSPEC spectrometer at the ESO 3.5-m NTT, at La Silla, Chile. We
concentrated on the H2 S(1) quadrupole line at 2.12 microns and the H3
multiplet at 3.5 microns, which both occur at low pressure levels, very
high in the stratosphere. During the week of impacts the 4."4 slit was
aligned along the parallel of the impact sites. This allowed to search
into the immediate implications of each impact to the Jovian atmosphere
at the time of event and to study the evolution of the older impact
sites additionally present in the spectra. Fragments B, F, and H were
observed at time of impact. The impact of B was observed at 2.12 microns
and H was watched in the 3.5 micron region. For details on the H event
see Encrenaz et al., this issue. For impact F we recorded data in 3
spectral regions (3.3 microns, 3.5 microns, 2.1 microns) within two
hours after impact. Since F almost fell on impact site E, both sites
could only be distinguished by their distinctly different spectra in the
2.1 micron region. The spectrum of F was characterized by a strong
featureless continuum, whereas the already evolved site E showed little
continuum, but a strong H2 line. After July 22, emissions clearly
associated to the impacts were detected in the 2.1 micron and 3.5 micron
regions also in the northern hemisphere (lat.: +44 degrees). The
emissions were not uniformly distributed in longitude, but concentrated
at the longitudes of the brightest impact sites. They could result from
the transfer of charged particles along the Jovian magnetic field lines.
The mapping of the entire planet in H2 and H(sub)3^+ showed their
distribution on the Jovian disk. As expected the H(sub)3^+ emission was
strongest in the auroral regions and was also strong at the impact sites
and their northern counterparts. Apart from these regions it was
extremely weak, although the strongest H(sub)3^+ line at 3.533 microns
could be detected at all latitudes.

05.08 Waite J. H., Jr.* Gladstone G. R. Na C. Stern S. A.
Franke K. Bagenal F. Clarke J. T. Fabian A. C. Brandt W. N.
X Ray Emissions Produced as a Result of the Impact of Comet
Shoemaker-Levy 9 with Jupiter

Rontgensatellit (ROSAT) high-resolution imager (HRI) observations of
Jupiter made before and during the impact of comet Shoemaker-Levy 9 with
Jupiter indicate the emission of x rays from northern mid-latitudes in
association with the K, P2, and W fragment impacts. The emissions
appear to originate at the north magnetic conjugate footprints
corresponding to southern latitude impact sites. This suggests a
mechanism whereby the comet impact produces in situ acceleration of
electrons and/or magnetospheric perturbations of trapped radiation belt
electrons that precipitate into the weaker surface magnetic field region
associated with the northern magnetic conjugate footpoint. Interaction
of the electrons with the upper atmosphere subsequently produces x ray
emissions through a bremsstrahlung process.

05.09 Dessler A. J.* Hill T. W.
Effects of Comet Impact on the Jovian Magnetosphere

As expected from observational estimates of dust concentration and upper
limits for gas production from Comet Shoemaker--Levy--9, direct effects
on the Jovian magnetosphere were slight [Dessler and Hill, GRL 21, 1043,
1994]. However, three unexpected effects of the comet impacts have been
reported: (1) approximately a 25% increase in the intensity of
decimetric radio emission (synchrotron radiation from trapped
relativistic electrons); (2) mid-latitude auroral UV emission features
in the northern hemisphere; and (3) x-ray emissions, also from mid-
latitudes in the northern hemisphere. The reported auroral features are
somewhat poleward of the northern footprint of the magnetic field line
that connects to the K impact site, according to available models of
Jupiter's intrinsic magnetic field, and the position of the source of
the x-ray emissions is consistent with that of the more highly resolved
auroral features. We propose that these three phenomena are different
manifestations of a single cause: namely, ionospheric motions driven by
the impacts of SL-9 fragments on the Jovian atmosphere. The impacts
produce ionospheric winds that interact with the Jovian magnetic field.
An ionospheric wind of 1 km/s produces a vxB electric field
approximately 0.5 V/m and a potential of more than 1 MV across the
region of impact-driven wind. The resulting magnetic-field-aligned
Birkeland current system accelerates electrons into the conjugate
atmosphere to produce the observed auroral arcs. The vxB electric f1eld
also "stirs" the magnetosphere near L = 2 (the region where decimetric
emissions are produced), which, because radio emission is sensitive to
radial magnetospheric motions, needs only a small displacement of the
relativistic electrons already present to account for the observed
decimetric enhancement. Thus, the one mechanism of splash-induced winds
explains the two auroral arcs and the associated x-ray feature in the
northern hemisphere as well as the increase in decimetric radio
emissions from the inner magnetosphere. This work was supported in part
by NSF grant ATM-9322360.

05.10-P DePoy D. L. Spencer J. R. Orton G. S. Dumas C. Kim S. J.
Near-IR Observations of the Comet Crash from Cerro Tololo

We used the 1-2.5 micrometer Ohio State Infrared Imager Spectrometer
(OSIRIS) to observe the Shoemaker-Levy 9 impacts from the CTIO 4-meter-
telescope in Chile. We obtained 1.58, 1.7, and 2.3 micrometer images and
1-2.5 micrometer cross-dispersed 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. In the CH4 filters, considerable structure was seen in
the high altitude clouds produced at many of the impact sites. The few
spectra of the impact site clouds so far reduced show weak K-band CH4
absorptions, probably due to CH4 above the impact clouds: simple
reflecting-layer 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

05.11-P Chanover N. J. Marley M. S. Severson S. A. Hereld M.
Near-Infrared Observations of Comet Shoemaker-Levy 9 Impacts

We collected near-infrared images of Jupiter with the Astrophysical
Research Consortium's 3.5 m telescope at Apache Point, NM, on the nights
of July 15, 16, 18, 19, and 20, 1994. The GRIM II near-infrared GRism
spectrometer/IMager was used to image Jupiter at several near-infrared
wavelengths with narrow-band filters centered at 1.58, 1.70, 1.99,
2.122, 2.22, and 2.36 microns and in K band. These data are used to
place limits on the vertical extents of the ejecta created by the comet
impacts. We compare the relative heights of the clouds created at
impact sites H, D/G, and E in an effort to characterize the ejecta
created by several different impacts. We also plan to integrate the APO
dataset with simultaneously obtained methane band and continuum images
from the Tortugas Mountain Observatory 0.6 m telescope, which span the
spectral region from 450 to 968 nm (see Kuehn et al., this issue). This
work was supported in part by NASA grant NAGW-1802.

05.12-P Kuehn D. M. Beebe R. F. Murrell A. S.
Vertical Structure Models for the Impact Regions on Jupiter of
Comet Shoemaker-Levy 9

The Tortugas Mountain Station of the NMSU Observatory continued its
monitoring of atmospheric changes on Jupiter throughout the 1994
observing season. Intensity calibrated images in broadband blue, green,
red, and near IR wavelengths as well as narrow band images of Jupiter in
visible and near-IR methane absorption bands were taken prior to, during
and after the Shoemaker-Levy 9 impacts in July 1994. Hubble Space
Telescope images obtained at similar wavelengths before, during and
after the impacts will be used as well. In particular, calibrated
ground-based observations were obtained for 20 July 1994 and 25 July

These high resolution (PSF FWHM approximately 1 arc sec for groundbased
imaging), photometrically calibrated images are used to retrieve
vertical structure information. Vertical structure models were
constructed by fitting scans of constant latitude (limb darkening with a
multilayer, radiative transfer model incorporating multiple scattering.
We are especially interested in the vertical structure of the localized
and resolved cloud disturbances arising from the comet impacts. The
scale of the time variation of any vertical structure detected will also
be of interest if it is possible to detect.

05.13-P Morozhenko A. V.
Use of Consequences of the Comet Shoemaker-Levy 9 Collision
with Jupiter for Study of Nature of Deep Atmospheric Layers of
Jupiter and of the Cometary Materials

Consequences of the collision which would be observed up to 1995 can be
changes of upper cloud layer of Jovian atmosphere and formation aerosol
haze of cometary origin. It is shown that this phenomena can be found
most effectively with simultaneous measurements of polarization and
reflectivity in several spectral ranges of continuum for wavelength from
0.30 to 0.80 microns and in contours of different methane absorption
bands. The measurements for the most powerful band at 0.889 microns make
it possible to find haze at the level of optical thickness near 0.02.

05.14-P Beebe R. F. Simon A. A. Huber L. F.
Comparison of Cloud Structure in Jupiter's Southern Hemisphere
at the Time of Shoemaker-Levy 9 Impacts with Voyager Data

Images with filter F336W (333+/-18 nm), F410M (409+/-07 nm), and F547M
(545+/-24 nm) or F555W (533+/-61 nm) obtained with the Hubble Space
Telescope Wide Field Planetary Camera 2 (WFPC2) are compared with 1979
Voyager 1 and 2 ultraviolet (325+/-45 nm), violet (400+/-50 nm) and
green (550+/-15 nm) images of similar spatial resolution. The extent of
alteration of typical cloud systems or generation of anomalous
structures associated with the comet is investigated. The expanded
wavelength coverage of WFPC2 (200-325 and 600-1000 nm) is utilized to
test the hypothesis that the dual morphology (turbulent spiraling aspect
or bright central region edged with a wavelike pattern) of cyclonic
structures at 28-44 degrees South latitude is due to differences in
stratospheric cloud cover. Support for this work is provided by NASA
through the SL-9 Campaign of the Space Telescope Science Institute,
which is operated by the Assoc. of Univ. for Research in Astronomy,
Inc., under contract NAS5-2G555.

05.15-P Yanamandra-Fisher P. A. Orton G. S. Gillam S. Young J. W.
Kanamori T. Spitale J. N. Grasdalen G. Fisher R. F.
Visual and Methane Band Observations of the Impact of Fragment
R on Jupiter from Table Mountain Observatory

Observations of Jupiter during the impact week of comet Shoemaker-Levy 9
were recorded at the JPL/Table Mountain Observatory (TMO) from 21 - 25
July 1994 (UT) using a visual CCD camera at several wavelengths: B and V
filters and four methane bands at 0.89 microns, 0.945 microns, 0.75
microns, 0.727 microns. The impact of fragment R was continuously
observed in the 0.89 micron methane band on 21 July 1994 from 5:00 to
6:00 UT at 50 second intervals. Observations of Jupiter in the other
filters were acquired pre- and post-R impact. We saw evidence for
several limb-brightening events between 5:30 UT and 5:45 UT, with the
impact site rotating onto the planet about 6:00 UT. The material appears
very bright at 0.89 microns, but is dark at all the other wavelengths,
indicating the deposition of material in the stratosphere. Current work
is focussed on the improvement of absolute calibration, determination of
center-to-limb variations at several wavelengths to constrain optical
properties of the stratospheric particles; evolution of the R impact
site and correlative analysis with near-infrared and thermal infrared
observations (from 1.65 microns - 20 microns) of Jupiter, acquired at
the NASA/InfraRed Telescope Facility during the same time interval.

05.16-P Musatenko S. I. Borovkov O. V. Ivchenko V. N. Kotsarenko A. N.
Possible Effects in Radio-frequency Emission of Jupiter on its
Interaction with Comet Shoemaker-Levy

Relying on study of radio-frequency emission of nearterrestial space in
ultra-short wave range and its connection with different
helioheophysical disturbances, the following effects would be expected
in radio-frequency emission during interaction process of magnetosphere,
ionosphere and atmosphere of Jupiter with the comet Shoemaker-Levy.

1. The radio-frequency emission fluxes bound to increase in the
centimeter and decimeter wave ranges.
2. The polarization degree in decimeter wave range will decrease and
next recover.
3. The radiobursts fluctuations dealing with orbiting of the planet
will decrease.
4. The discharges type of thunderstorm ones are occurred due to big
quantity of dust.

The implantation of the cometary fragments in the Jupiter atmosphere,
their motion and explosions will produce powerful shock and acoustic
waves. Propagation of these waves would cause fluctuations of
radiobursts in the centimeter wave range.

Our line of investigation will be consist in searching for
abovementioned effects.

05.17-P de Pater I. Heiles C. Wong M. Maddalena R. J. Strom R.
LePoole R. Spoelstra T. Bird M. Neidhoefer J. Funke O.
Price M. Kesteven M. Bolton S. J. Klein M. J. Galopeau P.
Gerard E. Lecacheux A. Dulk G. Leblanc Y. Hunstead R. W.
Campbell-Wilson D. Ye T.
Jupiter's Synchrotron Radiation Throughout the SL9 Impacts

Jupiter's microwave emission has been observed throughout the SL9 impact
period by many different telescopes. We will discuss the synchrotron
radiation data obtained with the 140-foot NRAO telescope in Green Bank
(20 cm), at Westerbork (20, 90 cm), Effelsberg (2.8, 6, and 11 cm),
Parkes (20 cm), NASA DSN (13 cm), Nancay (20, 18, 9 cm), the Australia
Telescope (13, 20 cm), the Molonglo Observatory Synthesis Telescope (36
cm) and the Very Large Array (20, 90 cm). The flux density increased by
roughly 20-30% during the week of impacts at most wavelengths. We will
present graphs of the total flux density as well as the spectral index
as a function of time at wavelengths between 2.8 and 90 cm. In
addition, we will show a few representive images, which emphasize the
brightness distribution and changes therein during the week of cometary

05.18-P Shemansky D. E. Matheson P. L. Festou M. A. Noll K. S.
Weaver H. A.
Upper Limits on the Amount of SL-9 Dust Contained in the Inner
Jovian Magnetosphere as Inferred from HST FOS Observations

Much of the dust and gas that accompanied the SL-9 comet into the Jovian
magnetosphere missed impacting Jupiter itself and continued through the
magnetosphere. Some material, particularly small dust grains, may have
become trapped in the magnetosphere when the qv x B force associated
with the grains charge was large compared to gravity. Horanyi, (GRL, 21,
1039-1042, 1994), has shown that the Io plasma torus is a likely
accumulation region for such grains. Molecular gases released in the
encounter processes have very short life times against dissociation and
ionization in the inner magnetosphere and can contribute plasma to the
magnetosphere also. The Hubble Space Telescope Faint Object Spectrograph
(FOS) has been used to observe the inner magnetosphere near the time of
impact of the SL-9 comet. We present results from these observations and
discuss limits on the amount of dust present in the inner magnetosphere.

A specific search for SiIII has been made because of this ions strong
2,1 transitions at 1892 Angstroms and because detection of silicon in
the torus would be unambiguously of cometary origin. CIII also has a
nearby line at 1910 Angstroms. Other strong SiIII transitions would be
effectively masked by OII, SIV, or hydrogen Lyman alpha radiation. The
nominal baseline criterion for detection by FOS at the 5 Rayleigh level
requires a column density of only 7 x 10^10 cm^-2. Silicon is sourced
from dust grains by ion impact sputtering and from energetic electron
induced dissociative reactions on grain surfaces. Estimated production
rates of silicon are low. The density of micron sized dust in the Io
plasma torus necessary to produce an observable effect is of order 10^-8
cm^-3. As of this writing, such dust densities are unlikely. If the
comet dust is ice-coated, as is likely at 5 AU, OH may be the dominant
product for energetic particle impacts. FOS observations searching for
OEI in the inner magnetosphere of Jupiter near the time of the impacts
may see some signal from diffuse cometary sources.

05.19-P Carsenty U. Mottola S. Braatz E. Neukum G.
Imaging of the Jovian Upper Atmosphere and the Io Torus During
the Encounter with Comet Shoemaker-Levy 9

A group of observers from the DLR institute for Planetary Exploration
(Berlin, Germany) will observe the encounter of the large fragments of
Comet Shoemaker-Levy 9 with Jupiter, using the 60cm Bochum telescope,
the DLR CCD Camera (lk x lk) and a set of narrow band interference

Imaging of Jupiter in the CH4 absorption bands (893nm,727nm) and
comparison with the "off" band (750nm), will enable us to monitor the
Jovian upper atmosphere and detect transient events as well as permanent
features, and study the interaction of the atmosphere with the cometary

Imaging of the Io plasma torus in the emission lines of NaI, [SII], and
[SIII] will enable us to study the interaction of the cometary fragments
with the Jovian magnetic field.

The nights before July 16 will be used to test our observing strategy
and to obtain a baseline of Jupiter data before the encounter.

We will implement an observing strategy which will enable us to observe
transient events as well as permanent effects of the encounter.

05.20-P McGrath M. A. Weaver H. A. Hall D. T. Schneider N. M.
Thomas N. Harris W. M. Livengood T. A. Prange R.
P/S-L9 Related Observations of the Io Plasma Torus

During the comet P/Shoemaker-Levy 9 encounter with the Jovian system
several sets of ultraviolet spectroscopic observations of the Io plasma
torus will be performed with both the Hubble Space Telescope and the
International Ultraviolet Explorer satellite. The scientific objectives
of these observing programs are two-fold: (1) search for new species,
particularly silicon and carbon, associated with the increased amounts
of dust being injected into the Jovian magnetosphere by the comet; and
(2) monitor the well-studied UV emissions from sulfur and oxygen ions
for any indication of significant change. The unambiguous detection of
either carbon or silicon would be indicative of comet remnants, since
the most sensitive UV spectroscopy of the torus to date with the Hopkins
Ultraviolet Telescope in December 1990 has failed to detect either, and
provides very stringent upper limits on their abundances. Variability by
as much as a factor of two in the line brightnesses on time scales of
weeks to months has been observed in both the UV and visible ion
emissions. Observed line brightnesses within the previously measured
range of variability will imply little impact by the comet on the
system, while variations outside this range may imply a fundamental
change such as that seen during the Pioneer 10 flyby in which the
"torus" was apparently only partial. We will report on the findings of
the planned observations.

05.21-P Hicks M. Grundy W. Fink U.
Spectral Imaging of the Io Plasma Torus and Spectrophotometry
of the G Impact Site

In March of 1994 we began a program of spectral observations of the Io
torus taken with our long-slit CCD spectrograph in a survey mode. By
moving the slit across the plane of sky between exposures we are able to
construct image cubes capturing all emissions between 0.5 and 1.0
micrometer and thus derive two-dimensional flux calibrated maps of
emissions of Na D (5890 angstrom), S^+ (67l6 angstrom) S^++ (9530
angstrom), etc. We first obtained baseline observations of the Io torus
before its interaction with P/Shoemaker-Levy 9, during the impact event
as well as post impact.

Our technique has the advantage over traditional filter imaging in that
we are able to observe a wide spectral range simultaneously and search
for emission changes which are not necessarily predicted. Weather
factors prevented us from obtaining emission maps in the midst of the
impact events. Some additional data was obtained Sept 6. At the time
that this abstract is being prepared we are only in the first stages of
data reduction. Preliminary analysis is showing time-variant phenomena.
Images of the torus at 5890 angstrom show an enhancement of the total
column density of Na D in the torus by a factor of three between April
14 UT and June 12 UT, unrelated to effects by the comet. This may be
correlated to volcanic activity on Io before or at this later date. It
may be possible that our studies can serve as an indirect monitor of
geologic activity.

In addition to studies of the torus, on July 22 UT, four days after its
collision with Jupiter, we observed the impact site of Fragment G with
our long slit spectrographic system. We present a calibrated spectrum
from 0.5 to 1.0 microns with a resolution of approximately l0 angstrom.
This spectrum should prove quite useful to help observers using
relatively wide methane filters do a proper job of continuum subtraction
on their images. When we compare the spectrum of the impact site with
the bright zonal material adjacent we see an approximately 10%
brightening in the saturated region of the 8900 angstrom methane band as
well as a gentle reddish slope at shorter wavelengths. We hope that our
spectrum will assist in constraining models of the impact sites, namely
the height of the aerosols and their optical depths.

This work was supported by NASA grant NAGW 1549.

05.22-P Niciejewski R. J. Roesler F. L. Scherb F. Oliversen R. J.
Observations of the Extended Sodium Cloud in the Jovian System
During the Encounter with Comet Shoemaker-Levy 9

High resolution spectrometric measurements of the sodium D emission rate
in the vicinity of Jupiter were conducted prior to, during, and
following the encounter with Comet Shoemaker-Levy 9. This paper will
present preliminary results from the measurement campaign and comment on
the magnitude of any change in the sodium magneto-nebula (zeno-corona).

05.23-P Prange R. Emerich C. Rego D. Ballester G. E. Clarke J. T.
Harris W. M. Livengood T. Zarka P. Ben Jaffel L. Beust H.
Dougherty M. K. Southwood D. Miller S. Gerard J. C. Ip W.
Horanyi M. McGrath M. Paresce F. Talavera A. Trauger J.
Staplefeldt K. Crisp D.
Auroral Signature of the Interaction of Comet Shoemaker-Levy 9
with the Jovian Magnetosphere

Observations of the FUV auroral emission of Jupiter have been performed
during the traversal of the Jovian magnetosphere by comet SL9 fragments,
together with an extensive set of preimpact (baseline) and postimpact
(relaxation) observations, using HST image complemented by IUE
observations. As part of this effort, we anticipate to identify the
interaction of the comet dust coma with magnetospheric plasma in
specific configurations where conjugate south/north auroral signatures
are expected using FOC images before and after comet in magnetosphere,
plus images looking for auroral arcs associated with P fragment.

05.24-P Budzien S. A. Livengood T. A. Fireman G. F. Harris W. M.
Ballester G. E. McGrath M. A. Prange R. M. Talavera A.
Early Results from the IUE Shoemaker-Levy Observing Campaign:
Temporal and Spatial Variability of the Jovian Ultraviolet

A comprehensive observation program of Jupiter was successfully executed
by the International Ultraviolet Explorer spacecraft SL9 science team.
Observations were conducted before, during, and after the impacts of the
comet Shoemaker-Levy 9 fragments with the Jovian atmosphere. These
impact events provided a unique opportunity to investigate the response
of the Jovian aurora to impulsive forcing on the magnetosphere. Relevant
processes capable of altering the auroral brightness or morphology may
include changes in ionospheric conductivity due to ion chemistry of
cometary volatiles or upwelling Jovian atmosphere, and modifications of
the co-rotation efficiency in the Io plasma torus due to loading by gas
and dust from the comet. The collisionally-excited far-ultraviolet H2
emissions from the north and south aurora were monitored, and no
dramatic changes in auroral morphology, brightness, or variability were
observed, although early indications are that the south auroral
intensity may have been somewhat lower than usual during the impact
period. Comparison with the typical temporal and spatial variability
will be made in light of more than 15 years of IUE observations of
Jupiter's aurora.

SAB is supported by a fellowship from the Office of Naval Research.

05.25-P Miller S. Achilleos N. Lam H. Tennyson J. Jagod M.-F.
Oka T. Geballe T. Joseph B. Baines K. Orton G.
Ballester G. Trafton L.
The Effect of the SL9 Impact on Jupiter's Aurorae

Spectra obtained on UKIRT and images from NASA's IRTF show that the
impact of Comet Shoemaker Levy-9 had a profound effect on the auroral
emissions of Jupiter as measured in infrared wavelengths sensitive to
the H(sub)3^+ molecular ion. In comparison with images and spectra taken
in previous years, the aurorae immediately prior to impact appeared
fairly normal. This situation continued at least until July 20;
observations immediately after the Fragment C impact (July 17) showed no
noticeable effect, for example.

But spectra taken on July 25 at UKIRT show a very considerable
enhancement of the northern aurora -- particulally in the region ,of
the maximum emission (circa cml 150 degrees). Preliminary analysis shows
that this is due to a tenfold increase in the column density of
H(sub)3^+, rather than an enhancement in temperature. In comparison with
previous years, the southern aurora appears slightly depressed.
Subsequent UKIRT spectra and images taken on the NASA IRTF show this
situation continuing at least until August 3. But later images show the
relative strengths of the north and south auroral zones returning to
pre-impact levels.

05.26-P Harris W. M. Ballester G. E. Barker J. Clarke J. Combi M.
Vincent M. Gladstone R. Kozyra J. Prange R. Bibring J.-P.
Emerich C. Ip W. Miller S. Rego D. Southwood D.
Talavera A. Dougherty M. Livengood T. A. Budzien S. A.
Espenak F. Fireman G. F. Kostiuk T. McGrath M. A.
Feldman P. D. Hall D. T. Strobel D. F. Moos H. W.
Woodney L. M.
Observing the Impact of Comet Shoemaker-Levy with Jupiter Using
the IUE Satellite

We describe our program to observe the effects of the impact of comet
Shoemaker-Levy with Jupiter using the IUE satellite. Four different
groups from the US and Europe combined their efforts into a single
science team with more than 450 hours of total satellite time between
June and September. The IUE spectral range from 1150-3300 angstrom
provided us with the most complete FUV/NUV coverage available during the
encounter, and the nearly continuous IUE duty cycle allowed direct
observation of all of the major impacts. During the course of this
program we were able to study comet related effects and to perform long
sought for simultaneous observations with the FOC, WFPC-II, EUVE, and
NSFCAM. The Jovian upper atmospheric and magnetospheric phenomena that
were monitored during this period include:

> The temporal evolution of the NUWFUV impact site albedos.
> The zonal and meridional migration of this new material by
stratospheric and thermospheric winds.
> The Jovian auroral emission intensity, optical depth, and relative
strength between the north and south zones.
> The high altitude equatorial Ly-alpha enhancement.
> The visible intensity of the Galilean satellites during the impacts
using the IUE targeting camera to search for evidence of a reflected
> The characteristics of the lo plasma torus.
> FUV emission from the developing impact plumes and their magnetically
conjugate regions in the northern hemisphere.

05.27-P Nazarchuk H. K. Shulman L. M.
The Preliminary Results of Spectral Monitoring of Jupiter and
Its Satellites During the Comet Impacts

The spectra of the Jovian satellites and the auroral zone of Jupiter are
obtained at 6-meter telescope to look for the Balmer and other emissions
expected in the impact phenomena.

05.28-P Lam H. A. Miller S. Tennyson J. Geballe T. R. Trafton L.
Condition Mapping of the Jovian Ionosphere Using H3+ as a
Probe--Baseline for the SL9 Impact Observation

Medium resolution spectra were taken in April 1993 using CGS4 mounted on
UKIRT. The data were fitted with theoretical H^+(sub)3 spectra to obtain
temperatures and column densities. The results show that the H^+(sub)3
auroral temeratures were higher than those on the body of the planet.
H^+(sub)3 emission is found at all latitudes with large enhancement at
the poles. Earlier work reporting elevated temperature around 102
degrees Central Meridian Longitude (CML) have not been repeated here.

Spectra taken with the slit along the equator detected H^+(sub)3
emission right the way across the planet. Taking into account the line
of sight effect it was found that the setting limb has a higher
concentration of H^+(sub)3 than at other CMLs.

This work will act as the baseline for the upcoming SL9 campaign .

05.29-P Gerasimov M. V. Dikov Yu. P. Yakovlev O. I. Wlotzka F.
Laboratory Approach to Impact of Meteorites and Comets on
Planetary Atmospheres

Early evolution of planets was characterized by high-temperature
processing of planetary matter in impacts during planetary accretion.
High-temperature processing of planetary matter and complex
heterogeneous chemistry during impacts provided production of multiple
chemical products which could be significant for initial global
evolution of a planet. The use of a powerful pulse laser is a reliable
method for laboratory simulation of high-temperature (3000-5000 K) pulse
processes such as hyper-velocity impact of large meteorite or a comet,
etc. In a cycle of our experimental works we have investigated chemical
processes which accompany high-temperature transformation of different
basic minerals, rocks, and meteorite matter.

Special attention was made for investigation of chemical effects of
counteraction of hot ejecta with planetary atmosphere. Obtained results
indicate extensive chemical processes involving hot ejecta and
atmospheric gases and can give a certain understanding of the early
evolutionary trends of primitive atmosphere, ocean and of silicate
matter. An approach is also done to simulate experimentally the
counteraction of a comet with reduced atmosphere.

05.30-P Gomez J. M. Lecacheux J. Colas F. Laques P. Rouan D.
Tiphene D. Sanchez-Lavega A.
Observations of the SL9 Impact on Jupiter from 0.4 to 2.16
Microns at Pic-du-Midi Observatory

We present the time-resolved photometric light curves at 2.16 microns
and films of the impacts of fragments H, L Q1 and S of comet SL9 with
Jupiter. At least three different phases are seen during the brightness
increase and up to its maximum of impacts H and L, whereas only a steady
increase is observed during the impact S. Photometric data in the nearby
bands at 1.25 microns (J) and 1.65 microns (H) are also given. We have
determined the spectral reflectivity from 0.4 to 2.16 microns (continuum
and methane bands) of cloud system G and L as measured on 20 July, as
well as their center-to-limb dependence. The tracking of the features in
the red since their formation and up to September, has allowed to
characterize their morphology changes and motions. The main spots
associated to each impact remained nearly stationary in System II (i.e.
their zonal velocities relative to the internal rotation perior --
System III -- were westward with u = -1 to -4 m/s). However the most
remarkable features (complexes G/D/S/R and K/W, and spot L) showed a
zonal expansion with velocities in the range u = 1 m/s to 31 m/s.
Besides this, meridional motions have been detected within several
impact regions, in particular within the complex G/D/S/R we have
measured meridional velocities v = 6 m/s, both in the northward and
southward directions from the impact center, reaching the cloud elements
on 1 August the latitudes 61 deg South (bright SPR limit in the 890 nm
methane band) and 37 deg S (STB).

05.31-P Scotti J. Larson S. Howell E. Marcialis R. Nolan M.
Wisniewski W. A'Hearn M. Wellnitz D. Campins H.
Rabinowitz D. Butner H. Tapia S. Vilas F. Doherty P.
Bessell M. Brosch N. Liebowitz E. Gilmore A. Kervin P.
Africano J. Levine S. Schuster W. Martin R. Verveer A.
Moreno F. Molina A. Rutten R. Kidger M.
Preliminary Results from CINE

CINE, experiencing about 50% clear skies during the SL9/Jupiter impact
week, amassed a large amount of data consisting of pre-impact images of
the comet fragments, spectra of satellites during impact, direct images
of the impact spots on Jupiter, and spectra of the spots. The comet had
become too faint for useful studies of the interaction with the Jovian
magnetic field, but the elongation of the inner coma by differential
acceleration was observed within 1-2 days of impact. Positions of
fragments E, G, L, P2 (8a and 8b),Ql,Q2, R, S, and W obtained at Las
Campanas on 1994 July 16 near 04:00 UT and reduced with respect to the
PPM astrometric catalogue generally indicate systematic residuals of -1"
to -2" RA with respect to the pre-impact ephemeris, consistent with the
impacts arriving approximately 10 minutes later than expected. With
instrumental dead time averaging 50%, flashes of the impact meteor off
salellites as continuum or spectral emissions were not observed at 2% of
the satellite's brightness. The evolution of the impact spots were
recorded with near continuous coverage during the impact period, and for
over two months at less regular inlervals. High resolution methane-band
images and high resolution long-slit spectra across the spots provide
information on the impact cloud height. High resolution narrow-band
methane images show the impact plume of H to be 750km above the limb
less than 20 min. after impact.

This project was funded by the NASA Planetary Astronomy Program, the
USAF Space Command and the various host institutions.

05.32-T Scherb F. Retherford K. Smyth W. H. Woodward R. C.
Comet Shoemaker-Levy 9 Impact with Jupiter: Observations of
[OI]6300 and Na D2 Emissions from Io

We plan to carry out observations of [OI]6300 and Na D(sub)2 (5890)
emissions near Io during a three-week period centered on the Comet
Shoemaker-Levy 9 impact with Jupiter. The observations will be carried
out with the stellar spectrograph at the McMath-Pierce telescope on Kitt
Peak. We will search for observable effects of the comet and its cloud
of dust and gas on the interaction of the plasma torus with Io's

This work was supported in part by NASA grant NAGW-3319.

SESSION 06 ....... Solar System Formation II
Monday, 4:00 - 5:30 Crystal Ballroom B
A. F. Berman and P. Cassen, Moderators

06.01 Boss A. P.* Yorke H. W.
Protoplanetary Disks, Mid-IR Dips, and Disk Gaps

The flat-topped spectral energy distributions of certain T Tauri stars
show evidence for significant dips at mid-infrared wavelengths (~10
micrometers) that could be explained by the existence of gaps in the
protoplanetary disks believed to be orbiting these pre-main-sequence
stars. Assuming an effective temperature depending on radius as T(sub)e
is proportional to r^-q, Adams, Lada, and Shu (1988, Ap.J., 326, 865)
showed that q ~ 0.5 disks lead to flat-topped spectra whose masses can
be inferred by fitting to long wavelength (~mm) observations. Marsh and
Mahoney (1992, Ap.J., 395, L115) showed that mid-IR dips could be caused
by removing disk matter with T(sub)e ~ 300K, corresponding to Lambda ~
10 micrometers (Wien's law). In this interpretation, the inferred disk
gaps could then be taken as evidence for unseen companions (possibly of
planetary mass) orbiting in the disk. Boss and Yorke (1993, Ap.J., 411,
L99) then showed that mid-IR dips could also be caused by the effects of
realistic dust grain opacities in a continuous disk. Using a detailed
radiative hydrodynamical model of a 0.02 solar mass protoplanetary disk
orbiting a solar-mass protostar (Boss 1993, Ap.J., 417, 351), b.y.
showed that the non-power-law T(sub)e profile produced by the
thermostatic action of the dust grain evaporation and condensation
process in a disk without a gap leads to a mid-IR dip quite similar to
that seen in the spectrum of T Tau itself. Using a refined technique for
calculating the radiative equilibrium of the disks, we have now
reproduced the initial result of b.y. for a 0.02 solar mass disk, and
extended the calculations to disks with masses of 0.01 solar mass and
0.04 solar mass. The disks are assumed to be accreting matter from their
envelopes (at ~10^-6-10^-5 solar mass/yr.) and hence are contracting
vertically, a process that appears to be important to disk energy
budgets; radial contraction occurs as well. The spectrum of the 0.04
solar mass disk shows a mid-IR hump (with a strong 10 micrometer
absorption feature) because in this higher mass disk, a region with
T(sub)e 250K extends to larger radii and so has a larger surface area.
The 0.01 solar mass disk shows an even more pronounced mid-IR dip than
the 0.02 solar mass disk because the T(sub)e profile contracts to
smaller radii in a lower mass disk. The 0.01 solar masses disk leads to
the best match to the spectrum of T Tau.

This research was partially supported by NASA grant NAGW-1410.

06.02 Graham J. A.* Boss A. P.
Clumpy Accretion During Early Stellar and Planetary Evolution

Three types of astronomical observations are best explained by the
existence of absorbing clumps of dust and gas that block stellar
radiation while moving at high velocities within a few astronomical
units of young pre-main sequence stars: (a) rapid variations in stellar
luminosity independent of wavelength; (b) rapid changes in surface
brightness of nearby reflection nebulae; and (c) rapid variations in
spectral lines excited in circumstellar regions. Clumps with mass of the
order of 10^22 grams appear to be required. Infalling motions
predominate. As the clumps impact a protoplanetary disk, a localized,
intense source of heat suitable for chondrule formation may occur. We
have been studying especially closely the pre-main sequence variable
star R CrA. New observations made this year at Las Campanas Observatory
provide evidence that the clumps exist over a wide range of opacities.
Optically thin clumps, which are responsible for circumstellar line
variations, are more common than the optically thick clouds necessary to
produce shadow effects on reflection nebulae. Multi-year AAVSO
observations of R CrA indicate that the number density of clouds changes
from year to year. While the origin of these clumps is presently
unknown, they might arise from the return of material swept up by the
outflowing stellar wind.

This work has been supported by NASA Grant No. NAGW-2287 in the Origins
of Solar Systems Research Program.

06.03 Tegler S.* Rettig T. Weintraub D. Pendleton Y. Whittet D.
Kulesa C.
Evidence for Chemical Processing of Pre-Cometary Icy Grains
Around Pre-Main-Sequence Stars

We report the detection of a broad absorption feature near 4.62
micrometers in the spectrum of the Taurus cloud source Elias 18. This
pre-main-sequence source is the second in Taurus, the third in our
survey, and the fifth known in the sky to show the broad 4.62
micrometers absorption feature. Of equal importance, this feature is not
seen toward several other embedded sources in our survey, nor is it seen
toward the source Elias 16, located behind the Taurus cloud. Laboratory
experiments with interstellar ice analogs show that such a feature is
associated with a complex C=N containing compound (called X(C=N)) that
results from high energy processing of simple ice components into more
complex, organic components.

We find a non-linear anticorrelation between the abundance of X(C=N) and
frozen CO in non-polar lattices. Because the abundances of frozen CO and
H20 are strongly correlated with each other and with visual extinction
toward sources embedded in and located behind the Taurus molecular
cloud, these ice components usually are associated with intracloud
material. Our results indicate that X(C=N) molecules result from
chemical processing of dust grains dominated by non-polar icy mantles in
the local environments of pre-main-sequence stars. Such processing of
icy grains in the early solar system may be an important source of
organic compounds observed in minor solar system bodies.

This research is supported by NASA grants NAGW-3433 and NAGW-3385 to
Notre Dame and Verderbilt Universitites.

06.04 Scholl H.* Lecavelier A. Roques F.
How To Detect a Planet in the Disk of Beta-Pictoris?

The Beta Pictoris disk exhibitsan inner clearing zone at a typical
distance of 15-30 A.U. from the star. We had investigated numerically
and analytically the possibility of dust confinement by a planet
orbiting Beta Pictoris (Icarus, 108, 37 and 59, 1994). Above a critical
planet mass of ~10^-5 stellar mass (5 Earth masses), the particles are
trapped in outer mean motion resonances. This mechanism can create a
steady state clearing zone extending inside the planet orbit. A
depletion region corotating with the planet and just outside it creates
an asymmetric feature in the dust disk. A moderate planet eccentricity
(lO^-2) can generate large arc-like structures.

We have improved this model, taking into account destructive collisions
between the particles in order to compare it directly to the
observations of the inner part of the disk. We show that, if the optical
depth is smaller than a critical value, assymetric structures created by
the planet are maintained in spite of collisions. These structures could
be a way to reveal bodies otherwise invisible by direct imagery.

06.05 Cochran W. D.* Hatzes A. P.
A Progress Report on the McDonald Observatory Search for
Extra-Solar Planetary Systems

The McDonald Observatory Planetary Search (MOPS) program is designed to
detect planetary companions to solar-type stars by measuring the orbital
motion of the star around the planetary system barycenter.We obtain
regular synoptic observations of the variations in stellar radial
velocity of a sample of 36 nearby F, G, and K dwarfs. Velocity
measurements are made to a precision of about 10m/s^-1, which would be
sufficient to detect a Jovian-mass planet around a solar-mass star. The
MOPS was started in September 1987, and now has accumulated seven years
worth of data. So far, no massive planets have been found with short
period orbits. We present typical results from the program, place limits
on the types of planets which can exist around program stars, and
discuss difficulties that have been encountered and their solutions. We
also present plans for future improvements in radial velocity precision
and survey sample size.

06.06 Cullers D. K.*
Statistical Model for Photometric Detection of Earth-Sized

A detector for Earth-sized planets orbiting other star which uses
photometry must see changes in apparen stellar luminosity of order
10^-4. Such changes, lastin several hours, occur at regularly spaced
intervals if cause by orbiting extra-solar planets intersecting the line
of sight to a distant star. Detection must occur against background of
random detector noise which causes fals positive events.

For highly plausible assumptions about causes and amounts of background
noise, an ideal folding detectio: algorithm finds Earth-sized planets
after 3-4 years of observing, at signal/noise ratios of about 3.5. The
noise model considers both instrument noise and that from fluctuations
in stellar spectra. These latter are assessed using detailed records of
solar luminosity fluctuations. It is found that the false alarm rate
curve is a steep function of detector threshold while planetary
detectability varies only slowly with this parameter.

This work is supported by NASA.

06.07 Jenkins J. M.* Doyle L. R.
Detection of Extrasolar Planets in Eclipsing Binary Systems: A
New Cross-Correlation Technique

The photometric detection of extrasolar planets in eclipsing binary
systems takes advantage of the binary orbital plane (and presumably the
planetary orbital plane) being parallel to the observer's line-of-sight.
A second advantage is the unique photometric signature a planetary
transit makes due to the changing configuration of the two stars,
allowing it to be distinguished from intrinsic stellar brightness
variations and other noise sources. The modeled convolution of planetary
transits with the light curve of selected eclipsing binaries can be
correlated with observational photometric data to retrieve planetary
transit signatures significantly below the noise. The usual ground-based
photometric limit of about 1% can therefore be exceeded. For the
smallest few eclipsing binary systems ground-based detection of
Earth-to-Neptune-sized planets should be possible with meter-class
telescopes. We have run a series of numerical experiments for the binary
system CM Draconis with twice Earth-sized planets in various circular
orbits. A matched filter is constructed that maxirnizes the
cross-correlation coefficient between the noise-corrupted observation
and the model. Although it is rather large, the parameter space provides
sufficient constraints to allow for unique detection of single planets
even when the transit drops are well below the standard deviation of the
measurement noise. In addition, it is possible to detect and isolate
transit signatures of multiple planets, if the observations are carried
out longer than two of the maximum orbital periods, and the planets are
well separated.

06.08 Gautier D.* Lecluse C. Robert F.
A Well-Mixed Solar Nebula: Constraints from Observed Deuterium
Abundances in the Solar System

Various objects of the Solar System are enhanced in deuterium with
respect to the protosolar abundance value of 2.6 x 10 ^-5 (Geiss, in
"Origin and Evolution of the Elements, Cambridge University Press, 1993
p. 89). This enrichment is frequently considered as resulting from
ion-molecules reactions which occurred in the interstellar medium prior
to the formation of the primitive solar nebula. This scenario implies
that the solar nebula was not well-mixed since otherwise the ices
enriched in deuterium would have lost their interstellar signature when
moving towards the warm inner part of the nebula. We show that D/H
ratios as high as 18 x 10 ^-5 can be reached in water in the nebula
through molecular reactions, when taking into account the deuterium
enrichment occuring at the gas-ice transition (Lecluse and Robert,
Geochem. Cosmochem. Acta, 1994, in press). Accordingly, the D-enrichment
observed in ocean water and in meteoritical clays as well as in the
atmospheres of Uranus, Neptune, Titan and the Halley comet may have
occurred in the nebula. This scenario is consistent with the wellmixed
nebula models advocated by a number of authors (Prinn and Fegley,
Astrophys. J, 249, 308, 1986, Prinn, Astrophys. J., 348, 725, 1990). On
the other hand, enrichments observed in hydrocarbons and organic
macromolecules embedded in carbonaceous chondrites have presumably an
interstellar origin and have never been subsequently reprocessed in the

06.09 Gustafson B. A. S.*
A New Analog to Light Scattering Laboratory at the University
of Florida

The absorption and scattering of sunlight by individual cosmic dust
grains and aerosols as well as the emission of infrared light can be
precisely simulated in the new scattering laboratory at the University
of Florida. Following the design of the now defunct microwave analog
facilities at the University of Florida and at the Ruhr University in
Germany, the new facility uses the principle of electromagnetic
similitude. Accurate measurements of single particle scattering and
precise control of the target parameters is achieved by magnifying the
wavelength and all particle dimensions by the same factor (~ 10^4) while
keeping the complex index of refraction of the material constant. Custom
designed antennas in a bistatic arrangement allow automatic filll
w-waveband coverage corresponding to the range from blue to red using a
network analyzer to record both scattered arnplitude and phase in the
angular interval from O to 165 degree scattering angle. The broad
wavelength coverage also allows the separation of the wanted signal from
the background using time-gating technique to increase accuracy and
boost the signal to noise ratio thus allowing scattering measurements
from particles close to the Rayleigh size limit. The upper size limit is
set by the extent of the flat portion of the wavefront which is of the
order of 10^2 wavelengths.
This work is supported by NASA's Planetary Atmospheres Program through
grant NAGW-2482.

06.10-P Lumme K. Rahola J.
Light Scattering by Porous Dust Aggregates in the
Discrete-Dipole Approximation

Electromagnetic scattering by various crystals and porous dust particles
has long been a challenging problem in several astronomical applications
including solar system objects as planetary atmospheres, cometary comae,
asteroid regoliths and interplanetary dust particles. In regoliths and
planetary atmospheres the particles strongly interact between each other
whereas in comae and interplanetary space we see effectively the single
scattering only.

Quite recently there has been a new computational progress in the
discrete-dipole approximation (DDA) shown by Lumme and Rahola (ApJ 425,
p. 633, 1994). We are now able to compute light scattering in almost any
geometry consisting up to 10^5 interacting elect dipoles. We are
currently studying stochastic aggregates which are thought to model dust
particles in the solar system. We show how the backscattering
enhancement and the reversal of linear polarization develop as a
function of the mean particle size, porosity and optical constants. Some
well known photometric and polarimetric features, such as the opposition
effect and the negative polarization seem to follow naturally from these

06.11-P Mautner M. Leonard R. L. Deamer D. W.
Meteorite Organics on the Early Earth: Hydrothermal Release,
Prebiotic Properties and Microbial Activity

The release, survival, prebiotic potential and microbial nutrient
properties of meteorite materials are examined experimentally. Simulated
undersea volcanic or impact-induced high temperature and pressure
conditions are applied to powdered Murchison meteorite, and release 10%
of the organics at 121 degrees C and 2.1 bars and 51% at 350 degrees C
and 250 bars, the latter including some of the organic polymer. The
hydrothermal release preserves prebiotic components such as carboxylic
acids, amino acids and polycyclic aromatics. The released material is
surface active, with surface pressures up to 2 x 10^-2 N/m and gives an
expanded surface tension isotherm that suggests a mixture of amphiphilic
components. One released component, nonanoic acid, forms vesicles.
Meteorite powder processed at high temperatures inhibits microbial
growth. However, the materials released at 121 degrees C are nutrients
for the humic acid utilizing microorganism Pseudomonas maltophilia, and
efficient nutrients for the oligotroph Flavobacterium oryzihabitans,
which grows to a globular shape in the dilute 10^-3 mol/l meteorite
extract medium and utilizes about 3% of the released organics. The
observations suggest that meteorite organics can be released efficiently
under early planetary conditions and are suitable for prebiotic
synthesis, and for supporting microorganisms on the early Earth or those
transported to other planetary environments. Hydrothermal and microbial
processing may be potentially useful in the direct conversion of
asteroid organics to biomass for space-based applications.

SESSION 07 ....... Educational Outreach Invited Posters
Monday, 5:30 - 6:30

07.01-P Buratti B. J. Brady L. E. Hanner M. S. Alvidrez R. F.
Roettger E. E.
`Teachers Touch the Sky:' A One Week Workshop for Teachers
Grades 4-6

Space science is inherently interesting to students, less threatening to
teachers than some other sciences, and interdisciplinary in nature.
These features make it the ideal vehicle for teaching basic scientific
concepts to children in a concrete and captivating manner. During the
summer of 1994, JPL scientists will be holding a one week pilot workshop
for 16 teachers in grades 4-6. We will walk the teachers through five
hands-on activities which are all based on current projects in astronomy
and space science at Caltech and the Jet Propulsion Lab. We will discuss
ways of integrating the lessons into the teachers' own curriculum; the
teaching of basic physical principals will be emphasized. The teachers
will also learn of the resources NASA makes available to them, and they
will have the opportunity to tour JPL's facilities and to talk to "real"
scientists about their work. The workshops to be covered include the
Venus topography box, cratering and mass,extinctions, and the
construction of a small telescope. The approach will be "small science"
in a supportive environment for teachers who in general have no
background in science.


07.02-P Cunningham C. C. Stiff T.
Curriculum-based Science Education Using a Multi-media Platform

A cooperative science education project with IBM, Canada, The Simcoe
County Board of Education (SCBE), and the Institute for Space and
Terrestrial Science (ISTS) has recently begun. The progress towards
developing a multi-media, Ontario high school curriculum based
demonstration product will be outlined. Concepts in space sciences will
be used to teach basic concepts in math and science to Ontario students
at the Grade 10 - OAC (Ontario Academic Credit) levels. The advantages
of this product over other multi-media learning materials currently
flooding the market will also be discussed.

07.03-P Limaye S. S. Sromovsky L. A. Krauss R. Wright E. Santek D.
Fry P. Saunders R. S.
McIDAS-eXplorer: A Software Environment for Analysis of
Planetary Data

A software environment for analysis of planetary data published by the
Planetary Data System on CD-ROM volumes has been developed for UNIX and
OS/2 2.1 operating systems. The environment allows convenient access,
navigation, analysis, display and animation of planetary data by
utilizing the full calibration and navigation data accompanying the
planetary data. Support currently exists for Voyager images of the giant
planets and their satellites, Magellan radar images (F-MIDR and C-MIDRs,
global map products (GxDRs) and altimetry data (ARCDRs), Galileo SSI
images of the earth, moon, Venus), Viking Mars images and MDIMs as well
as most earth based telescopic images of solar system objects (FITS).

The NAIF/JPL SPICE kernels are used for image navigation when available.
For data without the SPICE kelnels (such as the bulk of the Voyager
Jupiter and Saturn imagery and Pioneer Orbitel images of Venus), tools
based on NAIF toolkit allow the user to navigate the images
interactively. Multiple navigation types can be attached to a given
image (e.g for ring navigation and planet navigations in the same
image). Tools are available to perform common image processing tasks
such as digital filtering, cartographic mapping, map overlays, data
extraction. It is also possible to have different planetary radii for an
object such as Venus which requires a different radius for the surface
and for the cloud level.

A graphical user interface based on Tcl-Tk scripting language is
provided (UNIX only at present) for using the environment and also to
provide on-line help. It is possible for end users to add applications
of their own to the environment at any time.

This work was supported by USRA Contract# 5555-08 under NASA's Applied
Information Systems Research Program and by the Space Science and
Engineering Center at University of Wisconsin-Madison.

SESSION 08 ....... Mission Applications Posters
Monday, 5:30 - 6:30

08.01-P Borucki W. Koch D. Dunham E. Cullers D. Webster L.
Granados A. Ford C. Reitsema H. Cochran W. Bell J.
FRESIP: A Discovery Mission Concept to Find Earth-sized Planets
Around Solar Like Stars

The current nebular theory postulates that planets are a consequence of
the formation of stars from viscous accretion disks. Condensation from
the accretion disk favors the formation of small rocky planets in the
hot inner region, and the formation of gas giants in the cool outer
region. Consequently, terrestrial-type planets in inner orbits should be
commonplace (Wetherill, Science 253, 535, 1991). From geometrical
considerations, Borucki and Summers (Icarus 58, 121, 1984) have shown
that 1% of planetary systems resembling our solar system should show
transits for Earth-sized (or larger) planets. Thus a photometric
satellite that uses a wide field of view telescope and a large detector
array to simultaneously monitor 5000 target stars should detect 50
planetary systems. To differentiate regularly recurring transits from
statistical fluctuations of the stellar flux, one must observe over
several orbital periods so that the false positive rate can be reduced
to one event or less. A one-meter aperture telescope placed in a halo
orbit about either the Ll or L2 Lagrange points and viewing
perpendicular to both the orbital and ecliptic planes can view
continuously for the required period because neither the Sun, Earth, or
Moon would enter the fireld of view. Model calculations show that the
observations should provide statistically significant estimates of the
distributions of planetary size, orbital radius, coplanarity, and the
frequency of planetary systems that have Earth-sized planets in inner
orbits. Because approximately one half of the star systems observed will
be binary systems, the frequency of planetary systems orbiting either
one or both of the stars can also be determined.

08.02-P Hart H. M. Lubenow A. Wells E. N. Schultz A. B.
Using Hubble Space Telescope to Observe Solar System Objects

The HST pointing control system was designed to acquire and accurately
track any solar system target with an apparent motion on the sky less
than 0.21 arc-sec/second, including planets, planetary satellites,
planetary surface features, comets, and asteroids. The apparent motion
of a target is comprised of its geocentric motion and the parallax and
aberration due to the motion of HST in orbit about Earth. The problems
posed by the intrinsic motion of solar system targets are exaggerated by
the small field of view and high spatial and spectral resolution of the
instruments. In general, moving targets can be tracked to an accuracy
better than 0.05" for the duration of an observation. Other constraints
on the operation of the telescope include avoidance of bright objects
that can damage the instruments, visibility of the communications
satellites from HST, instrument operational constraints posed by the
South Atlantic Anomaly, restrictions on spacecraft pointing to allow
sufficient sunlight on the solar panels to maintain power, availability
of guide stars for the duration of the observation, etc. Most of these
constraints are invisible to an investigator proposing to use the HST,
but they all affect the final schedule of observations, and sometimes
can cause a particular observation to be unschedulable. This paper gives
an overview of the steps needed and systems used to prepare, schedule,
and execute observations of moving targets. The development state of new
capabilities is outlined. Some suggestions are given to help
investigators prepare to submit proposals.

The preparation of this poster was supported by Space Telescope Science

08.03-P Russell C. T. Metzger A. Pieters C. Elphic R. C. McCord T.
Head J. Abshire J. Phillips R. Sykes M. A'Hearn M.
Hickman M. Sercel J. Kluever C. Rosenthal R. Purdy W.
Maximizing the Scientific Return of Low Cost Planetary Missions
Using Solar Electric Propulsion

After many years of development solar electric propulsion is now a
practical low cost alternative for many planetary missions. The ion
thrusters and power processing units are now undergoing extended
operational tests at NASA Lewis Research Center and a space flight in
low earth orbit is being planned for the first flight of these
thrusters. Moreover, a large commercial market is considering the use of
these thrusters. Thus their availability and continued improvement seem

In response to the recent Discovery AO we and a number of colleagues
have examined the scientific return from a mission to map the moon and
then rendezvous with a small body. We have named this mission after
Diana, the Roman goddess of the moon and the hunt. In planning this
mission we found that solar electric propulsion was quite affordable
under the Discovery cost guidelines, that many targets could be reached
more rapidly with solar electric propulsion than chemical propulsion,
that a large number of planetary bodies were accessible with modest
propulsion systems, and that such missions were quite adaptable with
generous launch windows which minimized mission risks. Moreover, solar
electric propulsion is ideally suited for large payloads requiring a
large amount of power.

08.04-P Tedesco E. F. Price S. D. Paxton L. J. Walker R. G.
Planetary Astronomy with the Midcourse Space Experiment (MSX)

The Midcourse Space Experiment (MSX) satellite, funded and managed by
the Ballistic Missile Defense Organization (BMDO), will be a
long-duration, "observatory" style measurement platform that will
collect several terabytes of data on earth, earthlimb, and celestial
backgrounds. In particular, observations of the zodiacal background,
zodiacal dust bands, comet trails, asteroids, comets, Pluto, and targets
of opportunity are planned. MSX will be launched into a polar orbit in
late 1994. The spacecraft will carry a suite of state-of-the-art
sensors, including a cryogenic infrared scanning radiometer and
Fourier-transform spectrometer, several visible and ultraviolet imagers
and five UV and visual spectrographic imagers. The optical sensors cover
the spectrum from the far ultraviolet through the longwave infrared
(0.11 to 28 micrometers). While the principal focus of MSX is to collect
phenomenology data in support of ballistic missile defense objectives,
it will also be capable of collecting well-calibrated data in support of
a variety of civilian science objectives in earth and atmospheric remote
sensing and astronomy. This presentation describes some of the
instrumentation, discusses the planned observations, and outlines the
approach to reducing, analyzing, and archiving the large database that
will result from several years of data collection supporting a variety
of scientific objectives.

This work was supported under contract from the United States Air Force,
Phillips Laboratory, Geophysical Directorate and the Ballistic Missile
Defense Organization.

SESSION 09 ....... Uranus and Neptune
Tuesday, 8:30 - 10:00 Crystal Ballroom A
C. de Bergh and H. B. Hammel, Moderators

09.01 Rages K. A.* Pollack J. B.
Modeling Uranus' Temporal Variability

Analysis of spatially resolved Voyager images of Uranus has shown that
the subtle latitudinal banding seen on the planet is associated with
substantial variations in the optical depth and single scattering albedo
of the methane cloud layer above 1.3 bar [1]. Due to Uranus' extreme
axial tilt, this latitudinal variation in the atmospheric scattering
properties will account for some portion of the temporal variability
seen by Lockwood et al. [2].

Models of Uranus' scattering atmosphere have been derived for seven
latitudes between 0 degrees and 85 degrees S, using the spatially
resolved Voyager data. Assuming that the northern hemisphere is a mirror
image of the,southern hemisphere, changes in Uranus' aspect since 1g72
can account for only half the variation seen in the blue and 1/6 of the
variation seen in the yellow by Lockwood et al. during that time.
Varying only the optical depth and haze mixing fraction of the methane
cloud also fails to reproduce the observed spectral dependence of
Uranus' secular variation. It appears that some change in the wavelength
dependence of the stratospheric haze absorption coefficient (darkening
toward the red) will be necessary to fit the observations with a
hemispherically symmetric model atmosphere.

[1] Rages et al. (1991). Icarus 89, 359-367. [2] Lockwood et al. (1983).
Astrophys. J. 266, 402-414.

09.02 Lyons J. R.*
Constraints on Meteoroidal Influx Into the Upper Atmospheres of
Uranus and Neptune

Results from a recent analysis of meteoroid ablation rates in the
atmosphere of Neptune (Moses, 1992) have been coupled with photochemical
models of the upper atmospheres of Neptune and Uranus to yield
estirnates of stratospheric water profiles as a function of meteoroid
influx. Because water has never been detected in the upper atmospheres
of the giant planets, the tangential column opacities of the model water
profiles were compared with ultraviolet absorption measurements made by
Voyager to determine maximum water influxes. For Uranus an upper limit
of 4 x 10^6 water molecules cm^-2 sec^-l is obtained. For a Triton-like
meteoroid composition (30% water ice by mass), this flux is consistent
with an Oort-family particle population, but is not consistent with a
large population of planet-family dust particles. For Neptune the model
water profile is strongly dependent on the still uncertain eddy
diffusion coefficient, making it difficult to rule out a large
planet-family population of IDP's. However, an IDP population
sufficiently large to account for the CO observed in Neptune's
atmosphere can be ruled out.

09.03 Crisp D.* Trauger J. Stapelfeldt K. Brooke T. Clarke J.
Ballester G. Evans R. WFPC2 Science Team
Hubble Space Telescope Wide Field Planetary Camera 2
Observations of Neptune

The Hubble Space Telescope Wide Field Planetary Camera 2 was used to
observe Neptune on 28-29 June 1994. Planetary Camera images (0.046
arcsectpixel) of one hemisphere of the 2.3 arcsec disk were acquired in
7 spectral passbands (255, 300, 467, 588, 620, 673, and 890-nm). The
opposite hemisphere was observed in a subset of these passbands (300nm -
673 nm). The near-UV images show an almost featureless limb-darkened
disk There is also very little contrast at most visible wavelengths
outside of the strong methane bands, but the 673-nm images reveal
several features seen by Voyager 2^1, including the dark band near 60S
latitude. We do not see the Great Dark Spot, but it would not be obvious
if it were near the limb. Images in the strong methane bands at 619 and
890 nm show weak limb brightening, and several bright cloud features.
The northern hemisphere is occupied by a single, bright,
planet-encircling cloud band centered near 30N latitude. This may be the
same bright feature discovered last fall in ground-based images of
Neptune 2. The tropics also appear about 20% darker than the disk
average at these wavelengths. The southern hemisphere includes two
discrete bright features. The largest and brightest is centered at 30S
latitude, alld extends for least 40 degrees of longitude, like the
Bright Companion to the Great Dark Spot. There is also a thin cloud band
at 45S latitude, which almost encircles the planet. An atmospheric
radiative transfer model based on a multi-level, multi-stream discrete
ordinate method is being used to analyze the center-to-limb profiles in
these images to provide new constraints on the composition and vertical
structure of the radiatively- active gases and aerosols in Neptune's

References: 1) Smith et al. Science 246, 1422, (1989). 2) Harnmel el al.
Bull. Arner. Astron. Soc., 25,1077, (1993).

09.04 Lockwood G. W.* Hammel H. B. Mills J.
Variability of Neptune in 1994 Compared with 1989: Imaging,
Spectra, and Photometry

Neptune shows intrinsic variability on time scales ranging from days to
years. To investigate causes of such variation, we compare spectra and
photometry obtained in 1994 (contemporaneous with HST imaging) with 1989
ground-based and Voyager data. These include photoelectric photometry
(Lowell), spectrophotometry of the 6190-Angstrom methane band (Lowell),
near-infrared imaging and photometry at six wavelengths (Mauna Kea), and
Voyager imaging and photometry. Periodic short-term variations
correspond to the transit of the "Bright Companion" just south of the
"Great Dark Spot." We present Voyager images and photometric sequences
derived from them for several transit events, along with ground-based
photometry of the same events. Changes in the amplitude of the
variability is sometirnes due to changes in the Bright Companion and
sometimes due to other transient features rotating with slightly
different speeds.

Our 1994 data include CCD spectra from 5500 to 8100 Angstroms and
photoelectric photometry in visible light obtained on about a dozen
nights in May and June. Our analysis addresses the questions: Has the
amplitude of rotational modulation changed since 1989, which may
indicate changes in the size and distribution of the features causing
the variation? Do features detectable in images affect visible
wavelength spectra? Is the steadily rising brightness of Neptune seen
since 1985 associated with obvious changes in spectral features?

NASA and NSF supported this research under grants NAGW-2385 (HBH) and
ATM-9021100 (GWL), respectively.

09.05 Roques F.*
Evidence for Saturated Gravity Wave Spectra in the Upper
Atmospheres of Neptune and Titan from Ground-based Observations

We have analyzed systematically stellar occultations by Neptune
(observed at 0.89 and 2.2 micrometers between 1983 and 1991) and Titan
(occultation of 28 Sagittarii, July 3rd 1989). They exhibit strong
scintillation spikes, at pressure levels ranging from ~ 1 to 50
microbars. This scintillation translates, via an inversion technique,
into local temperature fluctuations of a few to several K, over vertical
distances of a fraction of a scale height, ~ 50 km (Hubbard et al.,
A.and A. 269, 541, 1993, Roques et al., A and A., 288, 985, 1994).

Although the time variations of these features are not available, their
spatial analysis can provide some clues as to their nature. First, the
gradient of temperature delta T/delta z is bounded from below by the
adiabatic lapse rate, i.e. ~ -1 K km^-l for both Neptune and Titan in
the regions under study. This appears to be true at all the scales
available from our observations, i.e. ~2-50 km in the vertical
direction. We note a strong asymmetry between the negative and positive
values of the gradient, the positive values being much steeper, reaching
values as high as ~+2-5 K km^-l.

Second, the power spectrum of the vertical temperature profiles, for
both Neptune and Titan, exhibit a characteristic power law variation in
m^-3 (where m is the vertical wavenumber) between vertical scales ,~ 2
and 50 km.

The two points above strongly suggest the existence of breaking internal
gravity waves in the upper stratospheres of both bodies. Further
analysis of existing occultation lightcurves by other planets (Jupiter,
Saturn, Uranus,...) is strongly recommended to look for similar

09.06 de Bergh C.* Lellouch E. Tokunaga A. Gautier D. Marten A.
Owen T. Rosenqvist J. Romani P.
Search for CO Absorptions in Near-Infrared Spectra of Neptune

Surprisingly large amounts of CO (mixing ratio of the order of 1O^-6)
have been detected in the stratosphere of Neptune from millimeter
measurements (Marten et al., Astrophys. J., vol.406, 285,1993;
Rosen-qvist et al., Astrophys. J., vol. 392, L99, 1992). Measurements in
the millimetric also showed that CO must be present in the troposphere
of Neptune in comparable amounts (Marten et al., ibid; Guilloteau et
al., Astron. Astrophys., vol. 279, 661, 1993; Naylor et al., Astron.
Astrophys. Lett., submitted). On the contrary, CO has never been
detected on Uranus. In August 1993, we obtained spectra of Neptune with
the CSHELL instrument at the NASA-IRTF telescope on Mauna Kea (Hawaii)
in order to look for absorptions of the 3-0 band of CO in the
troposphere of Neptune. The spectra cover the range 6328-6348 cm^-l at a
resolution of 20,000. Spectra of Uranus were recorded during the same
run in order to help eliminating the methane absorptions present in the
Neptune spectra. We find that the spectra of Neptune and Uranus at 1.6
micron look strikingly similar. Comparable methane absorptions are seen
in both planetary spectra, in addition to numerous telluric features.
There is no clear detection of CO absorptions in the spectrum of
Neptune. A comparison with results from the millimetric measurements is

09.07 DeBoer D. R.* Steffes P. G.
Radiative Transfer Results for Neptune Microwave Emission

Recently it has been shown that the microwave opacity of H2S is greater
than previously modeled and that it could possibly be responsible for
the microwave opacity in Neptune's troposphere measured by Voyager 2
radio occultation (DeBoer and Steffes Icarus 109, June 1994). This
paper examines that proposal in greater detail by comparing results from
a radiative transfer model with radio telescope and Voyager radio
occultation observations.

Sub-solar quantities of NH3 (< 5 ppm) are needed below the putative
NH4SH cloud near 35 bars to match the relatively high brightness
temperature near 1 GHz measured by the VLA (de Pater and Richmond,
Icarus 80, 1-13, 1989). This value (5 ppm) also reproduces the sharp
dip in emission measured above 1 GHz but requires that essentially no
H2S be present throughout the entire atmosphere. Other abundance
profiles which seem to better describe the measured spectrum between
100-200 GHz include relatively large H2S abundances along with NH3
depletions. These suffer at frequencies with weighting functions
peaking near the NH4SH cloud (roughly 5-10 GHz). The microwave
extinction due to NH4SH is currently poorly known and could possibly
play a role at these frequencies. Both profiles yield absorptivities
which are similar to those measured by Voyager 2 near 6 bars.

This work was supported by the NASA Planetary Atmospheres Program under
grant NAGW-533.

09.08-P Sromovsky L. A. Fry P. M. Hammel H. B. Spencer J. R.
Shure M.
Cloud Features in 1993 Images of Neptune at 1.6 and 2.2 Microns

Using the NSFCAM at NASA's Infrared Telescope Facility, 76 near IR
images of Neptune were acquired during a 2.5 hr period on 4 September
1993 (UT). Most were made with a 2.2 micron filter (K-band) at a scale
of approximately 0.056 arc seconds/pixel, for which Neptune's equatorial
diameter spans 41.4 pixels. In the best of these images the PSF had a
FWHM of 0.4 arc seconds. Although strong methane absorption makes
Neptune's atmosphere virtually invisible at 2.2 microns, discrete bright
features indicative of high altitude clouds were seen in all of the
images. Assigning latitudes to the observed features was complicated by
image smear and the lack of a visible limb profile, but was facilitated
by a star that appeared on each image within 10 arc seconds of Neptune.
The PSF determined from the star image was used to deconvolve Neptune's
image, greatly reducing the smear in some images. The motion of
Neptune's cloud features relative to the star was also used in a
non-linear regression algorithm to establish both the approximate pixel
scale and the approximate feature latitudes, under the assumption that
the features rotated about Neptune's axis at the same latitude-dependent
rate that was defined by the Voyager observations (Sromovsky et al.,
Icarus 105, 1993). Latitudes were also estimated from one image at 1.6
microns (H-band) for which an atmospheric limb could be identified.
Although the latitudes are still very preliminary, at least four
discrete cloud features are visible, at approximate planetocentric
latitudes of 55 degrees S, 40 degrees S, 10 degrees N and 25 degrees N.
The brightest feature was near 10 degrees N. For comparison, HST images
made during 1991 at 889 nm (Sromovsky et al., Bull. Am. Astron. Soc. 25,
1993) showed features at 66 degrees S, 50 degrees S, 6 degrees N, and
about 30-45 degrees N, all at locations where Voyager saw features
during 1989. Given the currently large uncertainties in the latitudes of
features in the IRTF images and very limited temporal coverage, as well
as spectral differences, it is not clear which of the 1993 features are
related to those seen earlier in 1989 Voyager images and 1991 HST
images. Imaging planned for 1994 will presumably clarify the nature of
these 1993 features.

09.09-P Pryor W. R. West R. A.
Aerosol Models of Neptune Applied to Voyager IRIS Radiometer

Neptune's radiative energy balance can be derived by combining data from
reflected sunlight with thermal emission data. We address the reflected
sunlight problem. Aerosol models have been derived from Voyager PPS data
[1], ground-based data [2], and Voyager camera data [3]. One additional
available constraint on Neptune aerosol models is broadband IRIS
radiometer data obtained at a variety of phase angles [4]. We apply the
PPS aerosol models (derived to fit data at 0.265 micrometers and 0.75
micrometers) to broadband IRIS radiometer data (0.3-1.9 micrometers) to
test their ability to reproduce the data. A comparison with other
aerosol models will be discussed. Each aerosol model can be used to make
predictions about the phase angle behavior of reflected sunlight in
Neptune's atmosphere. An aerosol model that successfully reproduces the
phase-angle behavior of the radiometer data can be used to obtain
improved estimates of Neptune's phase integral, and its radiative energy

This work was supported by NASA's Neptune Data Analysis Program.
1Pryor, W.R., R.A. West,, K.E. Simmons, and M. Delitsky, Icarus 99,
2Baines, K.H., and H. B. Hammel, submitted to Icarus, 1994.
3Moses, J.I., K. Rages, and J.B. Pollack, submitted to Icarus, 1994.
4Pearl, J.C., and B.J. Conrath, JGR 96,18921-18930,1991.

09.10-P Bishop J. Romani P. N. Atreya S. K.
Photochemical Modeling of Voyager UVS Solar Occultation
Lightcurves Acquired at Neptune

Ingress and egress occultations of the Sun by Neptune were recorded with
the UVS instrument on Voyager 2 [Broadfoot et al. 1989, Science
246:1459]. Previous modeling of the 125-140 nm lightcurves indicated
eddy mixing coefficient values of 3-10 x 10^6 cm^2 s^-1 near the 0.2
microbar level and methane mixing ratios in the lower stratosphere on
the order of 1-3 x 10^-4 [Bishop et al. 1992, J. Geophys. Res. Planets
97:11681]. These results should be insensitive to photochemical details,
provided methane is indeed the main source of opacity at these
wavelengths. The UVS lightcurves at longer wavelengths (140-153 nm) are
expected to be dominated by the opacity of C2 species (ethane,
acetylene, ethylene) and perhaps higher order organics. At still longer
wavelengths, H(sub)2 Rayleigh scattering is a major opacity source. We
will present photochemical models giving good agreement with both sets
of UVS l1ghtcurves at wavelengths 125-165 nm and with C(sub)2H(sub)6 and
C(sub)2H(sub)2 abundances near 0.5 mbar derived from IRIS measurements
[Bezard et al. 1991, J. Geophys. Res. 96:18961]. The current
photochemical model incorporates several updates, including the recent
revision in CH(sub)4 photolysis branching ratios at Lyman alpha
[Mordaunt et al. 1993, J. Chem. Phys. 98:2054]. These model fits
strongly suggest a reglon of enhanced eddy mixing near the 10 mubar
level (K > 108 cm2 s-l), decreasing at higher altitudes. Comparisons
between egress (~49 degrees S latitude) and ingress (61 degrees N) UVS
results will be discussed, as 'well as the impact of uncertainties in
mean stratospheric temperature and key reaction rates. In line with our
earlier work, methane mixing ratios on the order of l0^-4 are required
to obtain good agreement between the photochemical models and the UVS

This work supported by NASA Contract NASW-4806.

09.11-P Conrath B. J. Gierasch P. J. Smith M. D. Ustinov E. A.
Simultaneous Retrieval of Temperature and Para Hydrogen in
Neptune's Atmosphere

The possibility of the simultaneous retrieval of the temperature and the
para hydrogen fraction in the atmosphere of Neptune by inversion of
thermal emission spectra between 200 and 350 cm- 1 is investigated. The
sensitivity of the spectral radiance at each wavenumber to the
temperature and para hydrogen fraction at each atmospheric pressure
level is obtained through the calculation of appropriate functional
derivatives. The sensitivity of the radiance to the para hydrogen
profile at atmospheric levels between 250 mbar and 850 mbar is of
special interest; that in the spectral region between 200 and 245 cm^- 1
is anti-correlated with that between 245 and 300 c m ^- 1. In contrast,
the sensitivity of the radiance to the atmospheric temperature is
positively correlated in these spectral regions. This differing behavior
permits independent information on the two atmospheric parameters to be
obtained. An inversion algorithm for the simultaneous retrieval of
temperature and para hydrogen profiles has been applied to data from the
Voyager IRIS north/south mapping sequence. Small but apparently
significant departures of the para hydrogen fraction from the
thermodynamic equilibrium values are found. Implications of these
results for the thermodynamics and dynamics of Neptune's upper
troposphere are discussed. This work was supported in part by the NASA
Planetary Atmospheres Program.

09.12-P LeBeau R. P. Dowling T. E.
Simulations of Neptune's Atmospheric Dynamics

We present three-dimensional circulation simulations of the atmosphere
of Neptune using a global, primitive-equation model (the EPIC model)
with ten vertical layers. The Voyager wind data initialize the cloud-top
level, and various assumptions are made about the variation of wind
strength above and below the clouds, the standard model using a sech^2
(theta) vertical variation similar to Leovy and Orsolini (1993, Icarus
106, 392-405). Our primary goal is to constrain the three-dimensional
structure of Neptune's jet streams and vortices by forward modeling. The
atmospheric dynamics of Neptune combine aspects found in the atmosphere
of Jupiter and in that of Earth, providing an important bridge between
the jovian and terrestrial regimes. Like Jupiter, Neptune has a large
anticyclone, the Great Dark Spot, with an angular size and position
similar to that of the Great Red Spot. On the other hand, the zonal
winds of Neptune, Uranus and Earth each have a limited number of jets
and a retrograde equatorial flow, in contrast to the multiple-jets and
prograde equatorial flow found on Jupiter and Saturn. Our numerical
models indicate that Neptune's atmospheric dynamics may be more global
than on any other planet. We are investigating the role of
thermodynamics in shaping the zonal winds, including moist
thermodynamics and the latent heat effect of ortho-para hydrogen

09.13-P Mickelson M. E. Larson L. E.
The Temperature Dependent Absorption Coefficient for Methane at
5430 Angstroms and 6190 Angstroms

Improved quantitative values of the methane absorption coefficient in
the visible and near infrared spectral regions are of vital importance
in the modeling of the atmospheres of the outer planets and their
satellites. We have extended previous laboratory measurements of the
temperature dependent absorption coefficient of the visible methane
bands at 6190 Angstrom and made new measurements for the 5430 Angstrom
region. Spectra were obtained using a tunable dye laser system coupled
to a cooled 3 m base length White-type cell. The laser operates with a
line width of 0.016 cm^-l, and data were recorded at approximately one
Angstrom intervals. Absorption coefficients derived from these data were
averaged over 10 Angstrom intervals. The cooled cell was adjusted for an
optical path of 288 m for the 6190 Angstrom band and 582 m for the one
at 5430 Angstrom. These measurements were made at gas densities of 1.22,
0.74, and 0.25 amagats and at four temperatures down to 123K. The
results are compared with previous laboratory measurements. The
absorption coefficient was determined over a wavelength interval from
6000 to 6400 Angstrom and 5270 to 5600 Angstrom respectively for the two
bands. As an independent check on absolute intensities, for one
wavelength location in both pseudo-continuum regions of each band and at
each temperature, signal absorption intensity was measured as a function
of gas density by systematically freezing the methane out of the main
cell into a small container held at LN2 temperature.

This work was supported by NASA Planetary Atmospheres Program Grant
NAGW-1765 and by the W.M. Keck Foundation.

SESSION 10 ....... Clementine and Moons I
Tuesday, 8:30 - 10:00 Crystal Ballroom B
J. F. Bell, III and P. G. Lucey, Moderators

10.01 Pieters C. M.* Staid M. I. Fischer E. M. Tompkins S. He G.
Clementine Science Team
Products of the Impact Process: A Sharper View from Clementine

The BMDO Clementine spacecralt spent a little more than two months
orbiting the Moon during the spring of 1994 as part of an instrument and
spacecraft integration test. Digital images were acquired for the
surface using four co-aligned framing cameras sensitive to radiation
from the visible to mid-infrared, each with a different spatial
resolution and coverage. The NASA supported science team has been
analyzing and validating these data for scientific applications. Our
group has focused on the UVVIS camera which produced 5 channels of
multispectral data from 0.4 to 1.0 micrometers. Calibration issues
center on dark subtraction, flat fields, scattered light, frame transfer
signal, and frame-to-frame registration. First-order corrections have
been developed and initial multispectral data products have proven to be
of good quality and provide an excellent tool for general unit
definition of surface materials at a scale similar to Lunar Orbiter
(200-500 m). The five UVVIS spectral channels are sensitive to
variations of surface composition as well as to degree exposure
(maturity). Using preliminary calibrations, five channel image cube
mosaics have been produced for Giordano Bruno, Copemicus and Tycho.

Although both Copemicus and Tycho contain extensive deposits of impact
melt, the variety of material ejected from different depths have not
been mixed beyond recognition at either crater. A primary scale of
lithologic mixing at these two large craters appears to be on the order
of 1-2 km; for a given local terrain type large blocks of mafic-rich
material can be observed in close proximity to more feldspathic
material. Copernicus exhibits substantially more heterogeneity than
Tycho. The southern wall of Copernicus is particularly enriched in small
mafic-bearing zones. Many of these also exhibit a low albedo and suggest
a component of excavated basalt or dark mantle. The impact process does
not homogenize lithologies within the crater since different quadrants
of the wall at Copernicus remain distinct from neighboring sections. In
contrast to the diversity of materials observed at large impact craters
which probe deep into the crust, much of the spectral variation at the
small fresh crater Giordano Bruno on the farside is due to the varying
proportion of fresh material mixed with older weathering products (both
presumably feldspathic).

10.02 Jakosky B. M.* Lucey P. G.
Long-Wave Infrared (LWIR) Camera Observations of the Moon from

The Clementine spacecraft included a long-wave infrared camera to map
thermal emission from the lunar surface. The instrument uses HgCdTl
detectors in a 128 x 128 actively cooled focal-plane array. The field of
view is 1.0 x 1.0 degrees, corresponding to a resolution on the lunar
surface of about 55 m/pixel at periapsis. The passband is centered at
8.75 microns. Preliminary estimates of the performance in orbit at the
Moon indicate that the noise in the dark field observations is
approximately 1 DN. The signal on the day side varies between 0 and 100
DN; the 1 DN noise corresponds roughly to a noise-equivalent uncertainty
in the derived brightness temperature of 1 K at 400 K and 2 K at 300 K.
Observations will be used to map the dayside thermal emission.
Variations at low latitudes are due primarily to variations in the
albedo, while those at higher latitudes include the effects of
topography. In the polar regions, measurable emission comes from the
sunlit areas only. In addition, special sequences were performed to
observe sites of interest at multiple emission angles. The results will
be used to derive estimates of surface roughness at multiple scales and
of the thermal energy balance in the polar regions. In conjunction with
Apollo 17 Infrared Scanning Radiometer observations of nightside thermal
emission, the results will constrain particle size and the lunar thermal

10.03 Hawke B. R.* Peterson C. A. Blewett D. T. Budney C. J.
Coombs C. R. Lucey P. G. Smith G. A. Taylor G. J.
Remote Sensing and Geologic Studies of the Aristarchus Region
of the Moon

The region that includes the Aristarchus Plateau and Montes Harbinger is
one of the most geologically and compositionally diverse areas of the
lunar surface. In the course of a continuing spectral study of the
Aristarchus region, numerous near-IR (0.6- 2.5 micrometers) reflectance
spectra were obtained for a wide variety of surface units and analyzed
in order to extract compositional information. Both Earth-based and
Clementine multispectral images were also utilized. The dominant surface
unit is a regional dark mantle deposit (RDMD) of pyroclastic origin.
Spectra for several portions of the RDMD exhibit nearly identical
characteristics. These include steep infrared continua, low albedoes,
and very broad absorptions centered longward of 1 micrometers. The
spectra are indicative of uncrystallized Fe^2+-bearing glass. The RDMD
exhibits an extremely homogeneous composition and contamination seems to
be largely due to Aristarchus ejecta. Mare materials with various ages
and compositions are present on and around the Plateau. A spectrum
obtained for the wall of Vallis Schroteri clearly indicates the presence
of fresh mare basalt.

The spectra of highlands units in the Aristarchus region can be
classified into two distinct categories which reflect very different
mineralogies. One group is concentrated in the vicinity of Aristarchus
crater and includes most of the spectra of the interior and exterior
deposits of Aristarchus, the Cobra Head of Vallis Schroteri, and impact
craters which expose highlands debris from beneath the RDMD on the
Plateau. While this group can be divided into distinct subclasses, in
general it can be characterized as having shallow infrared continuum
slopes and fairly strong absorptions centered between 0.97 and 1.0
micrometers. These characteristics are indicative of a feldspar-bearing
material which has a mafic assemblage dominated by Ca-rich pyroxene. The
gabbroic composition of Imbrium ejecta exposed by small impact craters
on the Aristarchus Plateau is very different from the noritic material
emplaced by Imbrium southeast of the basin.

The other group of spectra were obtained for widely separated locations
in the region. These include Herodotus X, the southern rim crest of
Aristarchus, and portions of the Agricola and Harbinger Mountains. The
spectral characteristics of this group, chiefly the broad shallow
absorption centered at or beyond 1 micrometer, are attributed to large
amounts of olivine.

10.04 Duxbury T.* McEwen A. Eliason E. Edwards K. Smith D.
Davies M.
Clementine Global Geodetic/Cartographic Accuracy

The Clementine Mission provided complete global lunar imaging coverage,
which was augmented with LIDAR ranging data. The global imaging data,
tahen by the UVVis Camera, had a ground resolution of about 100-150
m/pixel while the LIDAR data had a ranging resolution of 45 m. UVVis
stereo imaging of the poles and other regions were obtained. HiRes
Camera imaging had an effective resolution of about 20m, but this was
not a global dataset. The Clementine spacecraft had two star trackers
for attitude control (instrument pointing) having an accuracy of about
0.03 deg (1 sigma). The spacecraft orbit (periselene altitude of ~400
km) was reconstructed to an accuracy of better than lOOm along track and
out-of-plane and better than 30m in range. Also, the instrument
alignments and geometric properties were determined to an accuracy of
better than 0.01 deg using star fields and locating the LIDAR footprint
in image coordinates.

Computing the lunar-fixed coordinates of any feature in a single UVVis
image would yield an accuracy of about 200m (1 sigma), anywhere on the
moon which represents a significant improvement in some areas having
uncertainties of over 5 km. Additional improvement for Clementine can be
made using standard photogrammetric techniques which takes advantage of
averaging the randomness of pointing and orbit errors over the global
datasets. A global cartographic control network can be produced having
an absolute accuracy of 50m or better in latitude and longitude. An
error in elevation, for those areas not covered by stereo imaging or by
LIDAR could be as large as 1 km since variations in topography of +/- 10
km have been seen with LIDAR. Good correlation was found to exist
between elevations derived from LIDAR and stereo imaging. Example of
cartographic accuracy including imaging and LIDAR derived topography
will be presented.

Parts of this research were performed at the California Institute of
Technology, Jet Propulsion Laboratory under the NASA Clementine Science
Team Program.

10.05 Cook A. C.* Roatsch T. Oberst J. Hoffmann H. Jaumann R.
Neukum G.
Semi-Automated Extraction of DTMs for Studies of the Lunar
Surface Using Apollo Metric and Clementine Stereo Pair Imagery

A technique is described that has been used to semi-automatically
generate Digital Terrain Models (DTMs) of the lunar surface using stereo
pair images obtained from the Apollo metric cameras, and from the
Clementine spacecraft. This involves using digital images of the Moon's
surface (scanned from photographs in the case of Metric) for automated
stereo matching with the "Gotcha" software developed at University
College London. The resulting stereo disparity ( data are then passed
through a camera model and, with the aid of some manually derived
control points, a DTM can be produced. This entire process can be
performed in under one day on a Sun Sparc 1 for a Metric stereo pair.

Apollo Metric stereo imagery has a ground resolution of 25m and a
height precision of 30m, but is limited in coverage to near equatorial
regions. Clementine stereo imagery has a ground pixel size of 100m at
best, and a height precision of the order of lkm over the majority of
the Moon. However in the polar areas, and other regions that were iamged
specifically for stereo, the height precision is improved to the order
of 100m.

The height information derived from the DTM yields important additional
data for the geological interpretation of surface features like: depth
to diameter ratio and volume studies for craters down to a diameter of l
km, the thickness of lava flows and crater ejecta, and cross-sections
through tectonic features and impact basin rings. In addition the DTM
allows us to enhance the morphology of surface features that normally
appear camouflaged by albedo markings or shading, and they can provide
local slope information to support lunar surface photometric studies.

Examples are given of a DTM covering the region to the west of
Tsiolkovskiy from Apollo Metric, and experiments to stereo polar regions
of the Moon from Clementine imagery. The goal of our future studies is
to produce regional scale DTMs and controlled rectified colour image
mosaics by combining Clementine data with other image data available for
the Moon.

10.06 Blewett D. T.* Hawke B. R. Lucey P. G. Spudis P. D.
Geologic and Compositional Studies of the Crisium Region of the
Moon Using Earth-Based and Spacecraft Spectra and Images

The Crisium basin is a multiring impact structure of Nectarian age
centered at 17.5 degrees N, 58.5 degrees E. The basin rim consists of
platform massifs, an unusual morphology possibly caused by post-impact
structural modification. Other interesting features in the vicinity of
Crisium include dark-halo impact craters, light plains units, probable
cryptomaria, and geochemical anomalies identified with data from Apollo
orbital X-ray and gamma-ray sensors. About 20 near-IR (0.6-2.5
micrometers) reflectance spectra for small areas in the Crisium region
were obtained with Univ. of Hawaii telescopes and instruments at Mauna
Kea Observatory (MKO). Spectral analysis focusing on the absorption band
near 1 micrometer provides information on Fe^+2-bearing minerals present
in the surfaces observed. In order to gain greater spatial coverage,
multispectral images in the extended visible (~0.4- 1.1 micrometer) have
been analyzed. The imagery includes that recorded by a CCD camera at
MKO, as well as data from the Galileo and Clementine spacecraft. Our
findings demonstrate that most circum-Crisium massifs are composed of
noritic anorthosite, though the slightly steeper continuum slopes in
some spectra with deeper bands may attest to the presence of
anorthositic norite. A spectrum for the central peak of the crater
Taruntius indicates the presence of a slightly more mafic (Fe,Mg-rich)
composition; this material may have been excavated from beneath the
Crisium ejecta deposit. Spectral and image mixing models show that
certain areas mapped as light plains and "smooth terra" NE of Taruntius
and in Palus Somni contain a significant component of mare basalt, i.e.,
cryptomaria are present. These areas have been identified as anomalous
in orbital geochemical data sets (e.g., Mg/Al greater than the
surrounding highlands). These units probably represent mare basalt
deposits contaminated with highlands debris by nearby impacts. Another
occurrence of smooth terra, to the SW of Crisium, also contains a large
mare component. However, this area has a rugged topography and is thus
unlikely to have been the site of a really extensive volcanic flows.
Hence we propose that an episode of explosive activity emplaced a mantle
of pyroclastic material in this area.

10.07 Bell J. F. III* Bregman J. D. Rank D. M. Temi P. Roush T. L.
Hawke B. R. Lucey P. G. Pollack J. B.
Mid-Infrared Imaging Spectroscopy of the Moon: 5.0 to 7.0
Micrometer Observations from the Kuiper Airborne Observatory

We have obtained the first mid-IR imaging spectroscopic observations of
the Moon. The data were obtained on the Kuiper Airborne Observatory
(KAO) using a 128x128 Si:Ga array camera mounted in a LHe dewar. Images
were obtained in October 1993 in 71 wavelengths from 5.0 to 7.0
micrometers using a cooled 1.5 percent resolution CVF filter. These
observations were exploratory in nature, and thus only one small (3.4
arcmin) region of the Moon was imaged. This study site is located along
the southwestern lunar limb near the craters Schickard, Baade, and
Inghirami. Although this region covers only 1 percent of the projected
area of the full Moon, it contains diverse geologic terrains, including
mare basalt/impact crater melt, highlands, and possibly buried
cryptonare. Thus it represents a useful test site for our preliminary

The 5 to 7 micrometers wavelength region represents a transitional area
of the spectrum between spectral regions dominated by reflected solar
flux (< 3 micrometers) and emitted thermal radiation (> 8 micrometers).
This wavelength range contains diagnostic mineral absorption features
that can be used to discriminate among many different potential lunar
rocks and minerals, including feldspar, olivine, pyroxene, and even
quartz. Because of the Earth's atmosphere, these wavelengths are not
accessible from groundbased telescopes. They can only be measured by
spacecraft or by the KAO (39,000 feet altitude).

Because of tbe exploratory nature of these observations, a rigorous
calibration of the data is not possible. However, we have devised a
relative calibration scheme that allows us to analyze the images for
evidence of compositional variations and to assess the detectability of
different rocks and minerals. This was accomplished by performing the
best possible offset and flatfield corrections to the images, spatially
coregistering the 71-channel image cube, and dividing out the average
spectrum of all lunar pixels in the cube. This provides the best
possible removal of instrumental and residual telluric artifacts in this
data set. The resulting images show spatially coherent evidence for
relative spectral variability at the +-10 percent level; some of this
variability correlates with albedo/morphologic features (mostly
craters), some does not.

Residual instrumental problems prevent the interpretation of spectral
features between 5.8 and 6.2 micrometers, but at other wavelengths, such
as near 5.6 micrometers and 6.7 micrometers, spectral features are
evident in ratio images and relative band depth maps. Spectral features
diagnostic of plagioclase and pyroxene exist at these wavelengths, but
continued refinement of the relative calibration is necessaly to
determine whether it is these minerals that we have actually detected
and mapped. At the very least it appears that the lunar surface is
detectably heterogeneous at these wavelengths, which is encouraging for
planned future, more detailed KAO and spacecraft mid-lR observations.

10.08 Henderson B. G.* Jakosky B. M.
Mid-IR Spectral Effects of Near-Surface Thermal Gradients in
Scattering Materials

Radiative cooling of the near-surface of evacuated particulate materials
can create significant thermal gradients in the uppermost layer where
the emitted energy is generated. A near-surface thermal gradient will
create emission features which are related to the spectral variation in
the opacity since it determines the depth from which the emitted energy
originates. We modeled this process for non-scattering materials and
showed that near-surface thermal gradients increase spectral contrast
and create emission maxima in the transparent regions of the spectrum
(Henderson and Jakosky, submitted to JGR). However, the mid-infrared
opacity of particulate materials is strongly affected by scattering as
well as absorption. It is thus important to investigate this problem for
scattering materials in order to interpret emission spectra of the
surronding surfaces of warm airless bodies like the Moon and Mercury.

We have examined the spectral effects of near-surface temperature
gradients on thermal emission spectra using Hapke's emittance theory and
numerical thermal gradient values from our previous model. Our results
show that near-surface thermal gradients enhance spectral contrast and
shift the wavelength location of the Christiansen peak to shorter
wavelengths, consistent with the laboratory data obtained by Logan et
al. (JGR, 78, 4983, 1973). This effect is due to increased scattered
contribution from warmer material at depth in regions of low absorption.
This result suggests that mid-IR spectral features due to a near-surface
thermal gradient are understandable and diagnostic of composition so
that mid-IR emission spectroscopy should be a useful tool for remote
determination of planetary surface composition.

10.09 Clark P. E.* Trombka J. I.
Multi-Variate Compositional Analysis Capabilities of Combined
Geochemical Experiment Package

The combined X-ray/Gamma-ray Experiment Package which flew on the Apollo
15 and 16 orbiters were provided data essential to the understanding to
our nearest neighbor (1). Specifically made available were elemental
compositional maps of the top (u (XRF) to cm (GRS)) of the lunar
regolith, an extensively gardened mixture of surface and buried rocks.
These compostionals maps show, by inference, the contribution of
underlying geochemical terraines and rock types found at the landing
sites. With these data (2,3,4,5), determinations of variations in
surface chemistry and major rock components have been made on local
(including landing sites and their surroundings), to regional (including
entire mare or ancient basins), to gobal (including the entire 10% to
20% area of data coverage) scales (e.g., 6,7,8). Multi-dimensional
classification techniques have been used to correlate XRF-derived Al and
Mg data, combined with GRS-derived Fe, Ti, and Th data. Two methods of
correlation, one supervised the other unsupervised, were developed. Such
mapping of relationships between variables is particularly important in
the case of Al and Mg, two elements with a complex relationship. One
correlation technique involved simply creating a density plot (two
dimensional array) from two dataset's values at a given pixel This 2D
array llas then placed into a 3 x 3 matrix consisting of 9 equal
sub-arrays containing an equal number of datapoints. The color assigned
to each sub-array was selected aneonic device, to allow an instant
estimate of the degree of positive or inverse correlation between the
two datasets at a pixel. The resulting color correlation map showed the
geographic distribution of the correlation. In this way, areas not
correlated in the normal way, anomalous in regard to expected rock
components near the surface, could be identified. The relationship
between Al/Si and albedo, elevation, and Fe were thus illustrated on a
global scale. Anomalies were found within the rings of Crisium,
surrounding sites of recent impacts and buried basins. The technique was
also used to indicate the nature of the distribution of Fra Hauro
basalts and ANT suite rocks in the Hadley Apennine region. The second
two-dimensional correlation technique developed was a supervised
classification, constrained by compositional fields of major rock types
occurring in the area of coverage. The population centers which appear
on a density plot of the two datasets are average soils consisting of a
mixture of major- components. In some cases, one major rock component
clearly dominates. Thus, such a plot is divided into units based both on
how population centers are distributed, and on how these fall relative
to the compositional fields of typical rock components. Population
centers are divided into two or more units to show how mixtures of
coeponents are distributed geographically in typical areas in a given
region. Such correlations were done in two highland regions, one in the
eastern highlands the other on the far side. This technique has also
been applied to global-scale correlations of Al/Si vs Fe,and Al\Si vs
Mg/Si. As a result, buried basalt deposits, indicated by moderately
sized areas with enhanced Ti, Ti, and/or Mg, have been shown to be
widespread in near and farside highlands. Unsupervised cluster analysis
was performed on all five datasets. The maximum likelihood classifier
consisted of an algorithm which combined parallelpiped and Bayesian
techniques. The algorithm was constrained to fit the data into the ten
most significant classes. The resulting regional units included three
mare units with correlation between basin age and Ti, the Hadley
Apennine area the Undarum/Spumans area, and five highland areas with
varying mafic components, most of which have been previously associated
with highland volcanism.

REFERENCES: Adler and Trombka (1970) Geochemical Exploration of the
Planets; [2] Arnold et al (1977) Proc Lun Sci Conf 8th 945-948; [3]
Bielefeld et al (1977) Proc Lun Sci Conf 8th, 901-908 [4] Hetzger et al
(1977 Proc Lun Sci Conf 8TH, 949-999; [5] Davis (1980) JGR 85 3209-3224;
[6] Clark et al (1981) Proc Lun Plan Sci Conf 12TH, 727-719; (7) Clark
et al (1978 Proc Lun Plan Sci Conf 9th, 3015-3027; (8) Clark et a1
(1991) Earth Moon Plan 53, 93-107.

10.10-P Simpson R. A. Lichtenberg L. Asmar S. W.
Clementine Bistatic Radar Experiment

Two sets of bistatic radar experiments were conducted while Clementine
(DSPSE) was in lunar orbit. In both cases, the spacecraft high-gain
antenna was aimed toward a surface target region and the scattered
signals were received at stations of the NASA Deep Space Network on
Earth. The transmitted signals were right circularly polarized at
Lambda=13 cm; both circular polarizations were received and recorded on
the ground. On 9-10 April 1994 experiments were conducted in a
"spotlight" mode over the south pole; on 23-24 April 1994 spotlight
experiments were conducted over the north pole. Part of one
quasi-specular track in the southem hemisphere was also obtained on the
second set of observations. The primary objective was to investigate the
scattering of radio waves in the polar regions, searching particularly
for signatures characteristic of ice such as have been reported on Mars
by Muhleman et al. (Science, 253, 1508-1513, 1991). Dates and times for
the observations were chosen so that the experimental geometry would
include the near-backscatter condition, allowing possible detection and
measurement of coherent backscatter effects (Hapke, Icarus, 88, 407-417,
1990). Early examination of the data shows no strong ice-like response,
in agreement with the results of Stacy (Cornell University, Ph.D.
dissertation, 1993) who studied high-resolution Arecibo radar images of
polar regions. But a broad enhancement of the scattering, such as may be
inferred for Mercury's polar regions (Harmon et al., Nature, 369,
213-215, 1994), is not ruled out. Detail in the echo spectra can be
matched with large features in Clementine mosaics and suggests that
radar photometric functions and/or low-resolution radar images may be
derived from these bistatic data with further analysis.

10.11-P Budney C. J. Lucey P. G.
Structure of the Mare in Southwestern Oceanus Procellarum

The total volume of mare basalts and the volumes of individual mare
units are important parameters for constraining the history and extent
of melting within the lunar mantle. How the mantle melted constrains
models of the bulk composition and thermal history of the Moon. Mare
basalt units have been spectrally mapped at low resolution from Earth
and the Galileo spacecraft. The Clementine mission has provided high
resolution multispectral digital data of most of the lunar surface. We
intend to use these data to map and estimate the volumes of highland
material and mare basalts of different compositions ejected from craters
in southwestern Oceanus Procellarum. We will use these data to infer the
depth and vertical structure of the mare, providing constraints on the
nature, timing, and production efficiency of mare volcanism in the
region. This work was supported by NASA grant NAGW1421.

SESSION 11 ....... Jupiter I
Tuesday, 10:30 - 12:00 Crystal Ballroom A
T. A. Livengood and R. Halthore, Moderators

11.01 Thomas A.* Prange R. Ballester G. E. Harris W. M.
Livengood T. Maurice S. Gerard J. C. Paresce F.
Comparison of IUE and HST Diagnostic of the Jovian Aurorae

Observation of FUV auroral emission from Jupiter have been performed for
more than a decade using the earth-orbiting ultraviolet observatory IUE.
One of the most striking characteristic of this emission is its
modulation in magnetic longitude, which has be interpreted as control of
auroral magnetospheric processes by the corotating asymetric magnetic

Recent HST images using the FOC has not confirmed this typical
modulation inferred from the IUE data. Using HST sequences of auroral
images in the Lyman H(sub)2 bands, we present simulations of IUE
response to the aurora. We show that they are consistent with IUE
observations, and that the longitudinal modulation is at least partly
due to geometrical effects in the IUE aperture. We discuss briefly
consequences of this finding.

11.02 Gladstone G. R.* Waite J. H. Jr. Na C. Y. Franke K.
Gerard J.-C. Grodent D.
Some Aeronomical Implications of Jupiter's Aurora

Recent observations of the jovian UV aurora with the HST Faint Object
Camera detected H2 band emissions of up to ~ 10 MR. Simulations of these
emissions suggest that the precipitating particle energy fluxes into the
auroral region must be routinely ~ 100 erg cm^-2 s^-l, and occasionally
as large as 1000 erg cm^-2 s^-l. For comparison, the available solar
energy flux at wavelengths < 1600 is only ~0.5 erg cm^-2 s^-l at
Jupiter. Such large energy inputs into a presumably low density region
of the atmosphere must dramatically disturb the upper atmosphere both in
the auroral regions and globally. UV auroras typically occur near the
homopause region, which, if low-latitude model atmospheres are used as a
guide, is about 350-400 km above the 1 bar pressure level. The pressure
and density in the region of the homopause are about 0.2-1 microbar and
6 x 10^l2-3 x 10^l3 cm^-3, respectively. Unless the auroral heating rate
(due to Joule heating as well as particle precipitation) nearly exactly
overlaps the cooling rate (presumably dominated by CH4 and C2H2 IR
radiation), then the expected exospheric temperatures are
extraordinarily large. Assuming steady-state, the exospheric temperature
expected for a column heating rate of 100 erg cm^-2 s^-1, with the peak
heating rate only 100 km above the peak cooling rate, is ~ 8000 K.

In this paper we present simulations of the FOC observations and
investigate some of the aeronomical implications of Jupiter's powerful
aurora. We also examine some other recent observations that seem to
require extremely large auroral-region exospheric temperatures.

11.03 Baron R. L.* Owen T. Connerney J. Satoh T. Harrington J.
Photometric Variability of the Jovian H3+ Aurora: A Strong
Correlation with Solar Wind Ram Pressure

Jupiter's 3.4 micron aurora was observed during the Ulysses spacecraft
encounter spanning the first months of 1992. After careful removal of
the consistent photometric changes due to Jovian rotation, the
integrated intensity of each polar region and the planet as a whole (a
quantity that changes markedly in tens of hours) show a strong
correlation with the solar wind ram pressure as measured by the Ulysses
spacecraft and projected to Jupiter. On the longer time scale of months,
photometric changes up to a factor of three in magnitude are observed. A
nearly quiescent state of the aurora appears to present itself just
after the Ulysses encounter.
This work was supported by NASA grant NASW 44 8 1

11.04 Satoh T.* Connerney J. E. P. Baron R. L. Owen T.
Observations and Modeling of Jovian H3+ Aurorae

Jovian H(sub)3^+ aurorae have been imaged (3.4 ,micrometers) on many
nights during the spring of 1992 using the ProtoCAM array camera on
NASA'S IRTF 3-m telescope at Mauna Kea, Hawaii. These high spatial
resolution (subarcsec) and high signal-to-noise ratio (several hundreds)
observations are compared with a distributed source model of the
aurorae. The source model uses auroral ovals corresponding to the last
closed field line (LCFL) computed using the GSFC O(sub)6 + current sheet
model magnetic field, but allows the position (longitude and latitude of
the oval center) and size of the ovals, the width, thickness, and height
of the emitting area to be adjusted. Due to the positioning
uncertainties in the planetary limbs (of the order of one pixel) and the
changing seeing condition through the observations, slight image shift
and different point-spread function (PSF) are also parameterized as well
as the source-model parameters. Several different source distributions
in the auroral region are examined as well as the uniform source
distribution model. Model parameters (including image shift and PSF) are
determined by utilizing the Generalized Inverse Method, so the
difference between the observed and model-computed brightness in
individual pixels of the images are minimized in a least square sense.
We conclude the source model based on the 0(sub)6-LCFL oval can be a
good first-order approximation to the Jovian H(sub)3+ aurora with the
above parameters adjusted properly. We also discuss the difference
between the northern and sourthern aurorae, as well as the short-term
variability of the auroral structure.

11.06 Ben Jaffel L.* Vidal-Madjar A. Clarke J. T. Emerich C.
Prange R. Gladstone R. McConnell J. Noll K.
HST-GHRS Detection of the Deuterium Lyman Alpha Emission at the
Limb of Jupiter

We have obtained high resolution profiles of the H and D Lyman alpha
lines emitted by the limb of the upper atmosphere of Jupiter, using the
G160M grating and the LSA of the GHRS instrument on board the Hubble
Space Telescope. The analysis of the resulting line shape provides,
through our radiative transfer calculations, the possiblity for the
first time to detect the deuterium Lyman Alpha emission at the planetary
limb. This observation represents a new observational technique giving
access to a possible new and independant evalution of the D/H ratio of
the proto-solar nebula as well as direct informations related to the
upper atmosphere physical parameters.

11.07 Emerich C.* Prange R. Ben Jaffel L. Clarke J. Ballester G.
Gladstone R. Sommeria J.
High Resolution Measurements of the H-Lyman Alpha Jovian Line

After the Goddard High Resolution Spectrograph (GHRS) side 1 of the
Hubble Space Telescope (HST) was repaired, high resolution measurements
of the Jovian H-Ly alpha line have been performed during the Cycle-4 of

Different regions, in both hemispheres as well as in the "bulge" itself,
have been selected and observed in order to spatially follow the
propagation of possible turbulent therrnospheric motions which could be
at the origin of the "bulge".

The first results of our analysis will be presented and discussed.

11.08 Karkoschka E.*
Spectrophotometry of the Jovian Planets and Titan: 300-1000 nm

Full-disk albedo spectra of the Jovian planets and Titan are presented.
They are based on observations taken in July, 1993 at the European Southern
Observatory. The accuracy is 2 percent for relative and 4 percent for
absolute albedos. The signal-to-noise ratio is approximately 1000 in the
visible. Ultraviolet albedos are 10-15 percent higher than many published
ones. UBV magnitudes were determined. They show some small temporal
variations with respect to published magnitudes, mostly in the ultraviolet.

The spectrum of each Jovian planet displays some 40-60 Raman scattering
features. They can be explained by a Raman scattering model with five
parameters adjusted for each planet. These parameters can be used in future
radiative transfer calculations.

A cold-temperature methane absorption spectrum has been determined which
fits the observed methane features. Care has been taken to separate
methane, ammonia, and hydrogen absorptions and Raman scattering features
which superpose at many wavelengths.

A strong absorption band at 933 nm wavelength in Jupiter's spectrum is
possibly due to water, confirming a strong depletion of oxygen in the
probed part of Jupiter's atmosphere.

11.09 Yair Y.* Levin Z. Tzivion S.
Detailed Evaluation of Cloud Growth, Charge Separation, and
Lightning Formation in Jupiter's Clouds

We use an axysimmetric cloud model with detailed microphysics to study
the formation and evolution of convective water clouds in the Jovian
troposphere, and to calculate the electrical structure of these clouds.
The model includes microphysical processes in both the liquid and ice
phases: nucleation, diffusional growth, two-phase coagulation, freezing
and sedimenation.

Charge separation is based on the non-inductive ice-ice mechanism, and
occurs due to interaction between graupel and ice particles in the
presence of supercooled liquid drops. We calculate the space charge
distributions and the electrical potentials and fields. Lightning is
assumed to occur wherever the field exceeds the breakdown value for the
ambient conditions.

Results show that the Jovian water clouds are very efficient in
separating electric charge. Electric fields reach values of 1.0-3.0
kV/cm as early as 30 minutes from the onset of ice formation in the
cloud. We calculate the flash frequency by assuming that each lightning
stroke neutralizes a 20 km channel within the cloud. The calculated rate
is several flashes per minute. and most discharges are located in the
upper part of the cloud at the 2 bar pressure level. The total energy
per flash, calculated from the reduction in the elecrical energy
density, is of the order of 10^l2J. Considering the optical efficiency
of Jupiter's atmosphere, the calculated optical energy for these
lightning flashes is 10^8-10^9J, in agreement with the analysis of
Voyager's data.

11.10-P Grodent D. Gerard J. C. Dols V.
The Spatial Distribution of the UV Color Ratio of the Jovian
Aurora Observed with the FOC/HST

Observations were made with the Faint Object Camera (FOC) on board
Hubble in July 1993 to investigate the time variation and the color
ratio distribution of the north UV Jovian aurora. Very bright discrete
emission (up to 4 MR) was observed in the H2 Lyman bands with a
considerable decrease 20 hours later. The morphology of the emission is
compared in 3 different UV bandpasses to map for the first time the
distribution of hydrocarbon absorption. Systematic differences are seen
between the H2 Wemer and Ly-alpha morphology on one side and the H2
Lyman on the other side. A tentative interpretation in terms of the
energy of the precipitating particles will be given.

If available at the time of the meeting, first FOC images of the UV
Saturn aurora will be briefly presented.

11.11-P Li Z. Hemtchinov V. Varanasi P.
Intensity, Line Width, and Line Shift Measurements in the
10-micrometer and Pure Rotation Bands of 14NH3 and 15NH3

The need for the line intensity, width, and shift data at low
temperatures of planetary interest on H(sub)2-, He-, and N(sub)2-
broadened lines in the 10 micrometers and pure rotation bands of
^l4NH(sub)3 and ^l5NH(sub)3 has prompted us to measure many of them
using a Fourier-transform spectrometer at several temperatures between
200 and 296 K. The variations of the line parameters with temperature
and quantum numbers are considered.

11.12-P Clapp M. L. Miller R. E.
Spectroscopy of Ammonia-Hydrazine and Ammonia-Water Mixed

Ammonia ice particles are thought to be the primary constituents of the
upper clouds of Jupiter and Saturn, yet no direct evidence of this has
been obtained from the infrared spectra of these planets. One
possibility is that condensation of ammonia on a foreign core may alter
the optical properties of the aerosol, masking its spectral features.
Hydrazine is believed to be a byproduct of ammonia photochemistry in the
Jovian stratosphere, and due to its low vapor pressure is a likely
candidate for the formation of condensation nuclei. In a low temperature
flow cell we have obtained spectra of ammonia coated hydrazine particles
from 700 to 4000 cm^-1 (14.3 to 2.5 micrometers). The infrared spectra
indicate that ammonia forms a crystalline coating on the hydrazine
particles. This composite system can be successfully modeled using Mie
theory and the published optical constants for ammonia. Crystalline
hydrazine optical constants were derived by a novel technique based on
aerosol extinction spectra which will also be described. The overall
conclusion is that ammonia-hydrazine interactions are too weak to
disrupt the ammonia spectral features.

The ammonia water system is much different. The strong interaction
between the ammonia and water molecules leads to efficient mixing of the
ammonia into the ice core, resulting in significant perturbations to the
ammonia spectrum. These spectra have been modeled using water ice
optical constants also derived in this laboratory.

11.13-P Borysow A. Zheng C.
Modeling of Collision-Induced Absorption Spectra of H2-H2 Pairs
in the First Overtone Band

A numerical method is presented [1] which generates the rotovibrational
collision-induced absorption (RV CIA) spectra of H2-H2 pairs in the
first overtone band of hydrogen, at temperatures from 20 to 500 K. The
first overtone band of the rotovibrational CIA of H2-H2 pairs, typically
referred to as the 1.2 micrometers, or (2-0) band, is associated with
two cases, a "single" vibrational transition upsilon(sub)l = O ---> 0,
upsilon(sub)2 = O ---> 2 (and similarly upsilon(sub)2 = 0 ---> 0,
upsilon(sub)1 = 0 ---> 1) and "double" vibrational transitions,
upsilon(sub)l = O ---> 1, upsilon(sub)2 = 0 ---> 1, where upsilon(sub)i
are the vibrational quantum numbers of two interacting H2 molecules (i =
1 or 2). The first overtone band appears in the near infrared region,
roughly between 7,500 and 10,000 cm^-l, which overlaps with the
frequency range of NIMS (Galileo) and VIMS (Cassini) future

The method employs spectral model lineshapes, whose parameters are
obtained from the lowest three, translational quantum mechanical
spectral moments. The model spectra reproduce very closely the results
of the quantum mechanical computations [2], which were based on the same
input. They also show good agreement with the existing experimental

The work is of interest for modeling the outer planets' atmospheres in
the near infrared region of the spectrum. User friendly FORTRAN program
utilizing the described method has been developed, and is available upon
request from the authors.

Acknowledgments: Partial support by NASA, Planetary Atmospheres Program,
grant NAGW-10-74507 is gratefully acknowledged.

[1] C. Zheng and A. Borysow, "Modeling of collision-induced infrared
absorption spectra of H2-H2 pairs in the first overtone band at
temperatures from 20 to 500 K", Icarus, 1994, submitted.
[2] W. Meyer and A. Borysow and L. Frommhold, "The binary
collision-induced first overtone band of gaseous hydrogen from first
principles", Phys. Rev. A 47, 1993, 4065-4077.

11.14-P Ustinov E. A.
Inverse Problems of Radiative Transfer and Inversion Methods of
Data Interpretations: A General Approach to Formulation and

The optical sounding of a planetary atmosphere, either remote or in
situ, may be regarded as an observation of a field of radiation produced
by the radiative transfer in the atmosphere. The nonlinear forward
operator converting relevant atmospheric parameters into the measured
radiative parameters is implemented by the radiative transfer
computations with subsequent convolution with characteristics of the
instrument. A general approach to the development of inversion methods
of interpretation of data of optical sounding is discussed that is based
on the formulation of the corresponding inverse problem of radiative
transfer. This inverse problem is considered as an implementation of an
inverse operator converting the measured data into the fields of the
atmospheric parameters to be retrieved. Three cases are considered:
explicit inversion of the forward operator (Case A), analytic
expressions for the weighting functions through the atmospheric
parameters (Case B), and, in a most general situation, these expressions
involve the solutions of forward and adjoint problems of radiative
transfer (Case C). Specific examples illustrating all three cases are

11.15-P Garcia-Melendo E. Sanchez-Lavega A. Gomez J. M. Lecacheux J.
Colas F. Parker D. Miyazaki I.
Study of a New Form of Activity Within the Jovian North
Temperate Eastward Jet

Since the 1990 Disturbance in Jupiter's 24 deg N high-speed jet, the
North Temperate Belt has been a prominent low albedo feature. During the
period 1991-1994 a series of dark spots potruding into the NTrZ and
running along the south edge of the NTB (average latitude 23.5 deg. N)
have been tracked from more than 10000 CCD images taken at different
wavelengths in the visible region (300 nm - 1000 nm). Some of these
features are long enduring, persisting for two years or more,
representing a record of longevity for the spots observed within the
latitude of the NTB jet. Their velocity was in average 125 m/s although
an oscillatory movement is appreciable in a latitude-longitude diagram
with amplitude 1 deg. and spatial period 360 deg. The derived velocities
relative to System III indicate deviations relative to the mean Voyager
reference profile, which could be due to a real change in the profile,
or to local motions of the features (wave propagation) or to vertical
sheared motions since the features seems to be deep seated as they have
high emissivity at 5 microns (G. Orton, priv. comminication) .

11.16-P Sanchez-Lavega A. Lecacheux J. Colas F. Parker D.
Miyazaki I. Gomez J. M.
A New Long-lived Anticyclonic Vortex in the South Tropical Zone
of Jupiter

A conspicuous synoptic scale vortex is being observed since 1983 in the
South Tropical Zone of Jupiter. Its center has varied within the
latitude range 19 deg S - 23 deg S, i.e. close to that of the GRS. The
spot has showed an erratic motion with extreme zonal velocities of -3.5
m/s and 0 m/s relative to System III, and a changing color, being white
most of the time (period 1983-1992, and 1994) but red in 1993. Images
obtained under the 890 nm methane band in 1993 (red period) showed it to
have a high albedo, resembling the GRS. The vortex has survived to the
intense activity developed in the SEB-STrZ region following the SEB
Disturbances of 1990 and 1?93. High resolution images obtained at
Pic-du-Midi Observatory showed it to have a size of 8000-10000 km
(East-West) times 6000 km (North-South) with an spiral shape band within
it rotating anticyclonically. Because of its latitude location,
longevity and morphology, this vortex represents an important clue to
understand the nature of the GRS itself and other jovian vortices.

11.17-P Smith M. D. Gierasch P. J. Conrath B. J.
Convective Adjustment on the Outer Planets Including Ortho-Para
Hydrogen Conversion

Convective adjustment and diffusive schemes based on terrestrial methods
are extended to outer planet applications in order to develop an
algorithm that can be used in dynamical modeling of outer planetary
flows. Over the temperature range between 50 K and 250 K the delayed
conversion of ortho-hydrogen to parahydrogen can have large effects on
stratification. One dimensional modeling of atmospheric layers cooled
from the top, as by radiation to space, is reported. After ortho- and
para-hydrogen are mixed by a convective event, the ortho/para ratio
begins to relax toward thermodynanlic equilibrium. The energy release
during relaxation drives the atmosphere toward stability. On Jupiter,
the influence of hydrogen conversion on convection is small. On Neptune
the energy released by hydrogen conversion is large, and it shuts off
convection. A time dependent structure ensues, with irregular occurrence
of mixing events associated with temperature fiuctuations of 1 K or
more. The mean thermal structure at levels where the temperature is less
than about 250 K is significantly stable, in contrast to the predictions
of conventional outer planet radiative-convective calculations.

11.18-P Fisher B. M. Jones B. Hudson H. S.
Observations of Jovian Thermal Structure at Mid-Infrared

The mid-infrared wavelengths offer a relatively clear view of the
thennal structure of the Jovian troposphere and (as we confinn) some
infonnation about the interior. We have conducted time-series imaging
observations at 10 micrometers and 20 micrometers with a 16 X 64-element
Si : As photoconductive detector array, using the 1.5 m UCSD/Minnesota
telescope at Mt. Lemmon, Arizona (2776 m). The data have a basic
one-minute interval between images to make p-mode searches possible,
following the observing procedure pioneered by Deming et al. (1989). At
20 micrometers we obtained about 1200 good images during four clear
nights over the interval 23 Feb. - 27 Feb. 1992, our best observing run.
Each of these images has a signal-to-noise ratio of about two on the
thermal features, which themselves have low contrast and carmot easily
be seen in the raw data prior to flat-fielding and flattening of the
limb darkening across the Jovian disk. We find the following properties:
(a) At 20 micrometers the thermal features rotate at System III (the
interior rate), as previously shown by Magalhaes et al. (1990); at 10
micrometers the rotation shifts between System I and System III as a
function of latitude. (b) The spatial distribution of the features
follows the band structure in latitude and has a semi-regular appearance
in longitude. There is, however, no dominant preferred spatial
frequency, nor are the spectral properties the same in N and S
hemispheres. (c) Individual features evolve significantly in time over
the S-day interval of the observations. The thermal features presumably
represent the tropospheric signatures of deep-seated convective
structures, but these properties have not yet been explained
theoretically. This work was supported by NASA grant NGT-50776.

11.19-P Flasar F. M. Conrath B. J.
Thermal Waves in Jupiter's Atmosphere

We have reexamined the zonal structure of temperatures retrieved from
the IRIS north-south mapping sequences of Voyager 1 and 2. This
structure is dominated at several latitudes by the wavenumber-1
components in the upper troposphere that Magalhaes et al. (1990, Icarus,
88, 39-72) have determined to be quasi-stationary with respect to System
III. We find that the strongest amplitudes of these waves tend to be
located near nulls in the meridional profile of the zonal wind,
extending usually into the cores of the eastward currents. Remarkably,
the waves appear to extend coherently over latitude: We find little
difference in the zonal phase of these waves with respect to System III
over the latitude range 28 degrees N to 20 degrees S. Wavenumber-1
features identified beyond 25 degrees S have zonal phases that differ by
approximately 180 degrees ; no statistically significant wavenumber-1
features have been identified at mid or high northern latitudes. All
this suggests the possibility of a globally coherent wave field whose
spatial structure is determined by the thermal stratification and winds
of the mean atmosphere. We will present models of planetary waves to
illustrate the possible ducting and tunneling properties of these waves.

This work has been supported by the NASA/Planetary Atmospheres Program.

11.20-P Lederer S. M. Marley M. S. Mosser B. Chanover N. J. Beebe R.
The Sensitivity of Jovian Seismological Observations to Albedo

Schmider et al. (A&A 1991 248, 281) and Mosser et al. (A & A 1992 267,
604) report detection of Doppler shifts in reflected solar radiation
from Jupiter induced by p-mode oscillations of the planet. We have
considered the possibility that these observations may be recording
albedo features in Jupiter's atmosphere rather than or in addition to
Jovian oscillations. We have modeled the Fourier Transform Spestroscopy
method of p-mode detection in detail using .889 micrometers methane-band
images of Jupiter to test the albedo feature hypothesis. We found that
the low frequency portion of the spectrum is strongly contaminated by
the photometric signal of albedo features rotating with Jupiter. In
particular, the equatorial plumes evident between 3 degrees N and 11
degrees N latitude create a signal at 250,mu Hz which coincides with
signal apparent in the results of both Schmider et al. (1991) and Mosser
et al. (1992). These plumes are also responsible for power at
frequencies ranging from 180 mu Hz to 1070 ,mu Hz. Furthermore, features
including white ovals at 33 degrees S are a likely source of signals
ranging from 400 mu Hz to 1600 mu Hz. Although albedo features
apparently are not responsible for the bulk of the higher frequency
signal attributed to p-modes, they do appear to significantly affect the
lower frequency portion of the power spectrum obtained by Mosser et al.
Ideally, future observations of Jupiter will incorporate simultaneous
imaging of Jupiter, thus eliminating any such ambiguity.

11.21-P Vidmatchenko A. P.
Brightness Variations of Jupiter, Saturn, and Transparency
Variations of the Earth Atmosphere

An investigation of brightness variations of Jupiter, Saturn, and stars
by mathematical spectral analysis has been made. Oscillation periods of
the Earth T ~2.5, 3.6, 5.2, 7.1, 8.9, 11.2, 16.5, 20, 48, 104, and 208
min. with some variations from day to day are obtained. Most stable
values are T ~ 20 and 104 min. Periods of brightness oscillations for
Saturn (137 and 179 min.) and for Jupiter (142 min.) were obtained. We
assumed that the last ones are free oscillations of giant planet's

SESSION 12 ....... Moons II and Mercury
Tuesday, 10:30 - 12:00 Crystal Ballroom B
B. C. Flynn and T. L. Roush, Moderators

12.01 Wisdom J.* Touma J.
Evolution of the Earth-Moon System

The tidal evolution of the Earth-Moon system is reexamined. Several
models of tidal friction are first compared in an averaged Hamiltonian
formulation of the dynamics. With one of these models, full integrations
of the tidally evolving Earth-Moon system are carried out in the
complete, fully interacting, and chaotically evolving planetary system.
Classic results on the history of the lunar orbit are confirmed by our
more general model. A detailed history of the obliquity of the Earth
which takes into account the evolving lunar orbit is presented.

This work was supported in part by the NASA Planetary Geology and
Geophysics program under grant NAGW-706, and by an NSF Presidential
Young Investigator Award AST-8857365.

12.02 Hartung J. B.*
Chi Scorpiid, Theta Ophiuchid and Corvid Meteorites are
Giordano Bruno Ejecta

If the orbit of meteoroids from a particular stream and the orbit of
specific ejecta from a particular impact can be shown to be the same,
then it is reasonable to conclude that the stream meteoroids were
produced by the impact. The orbit of an object is specified by its
position and velocity at a given time. Both the above named meteor
streams and tne Giordano Brun impact occurred in June, so they are found
to have the same position at a given time.

The velocity magnitudes for stream meteoroids and impact ejecta are not
determined precisely enough for a useful comparison of orbits. The
velocity direction, or anti-radiant, of a meteor stream is specified by
its observed radiant right ascension and declination (equatorial
coordinates), which are generally known within a few degrees. Ejecta
directions are specified by their azimuth and elevation angles (nautical
coordinates). Ejecta elevation angles tend to be between 40 and 60
degrees, but can range widely during a single impact. Therefore, the
elevation angle is of little value as a basis for comparing similar

So, only the azimuth angle remains as a possible basis for establishing
the similarity of two orbits. The azimuth of meteoroid motion is
determined via a coordinate transformation, equatorial to nautical, and
is uncertain by only few degrees. To express both the meteoroid and
ejecta directions of motion in the same coordinate system requires
knowledge of the location of the crater on the Moon and the time of the
impact, both of which are accurately known for the crater, Giordano
Bruno (lunar lat. = 36 deg. N.; long. = 103 deg. E 2100 hr. GMT: June
19, 1178, Julian calendar).

Here an important assumption is made that the azimuths of the most
abundant ejecta from the impact are the same as the azimuths of the most
prominent rays extending from the crater. We find from the study of
photographs and maps of the Moon that the azimuths of the three most
prominent rays from Giordano Bruno to be 125 deg., 150 deg., and 237
deg. The azimuths, derived from observed radiants of the theta
Ophiuchid, Ch Scorpiid and Corvid meteoroids are 123 deg., 155 deg., and
238 deg. respectively. The remarkable similarity of these three pairs of
azimuths is beyond occurrence by chance and, therefore, provide a basis
for the conclusion that these meteor streams were produced by the
Giordano Bruno impact.

These results are based on an assumption of simple two-body motion; no
perturbations occurred to change the orbit parameters between the time
of the impact and the time the meteor streams were observed. If
perturbations did occur, which is likely, then additional streams, such
as the Librids, June Bootids, June Lyrids, and Tau Herculids also may
have been produced by the Giordano Bruno impact.

12.03 Gladman B. J.* Burns J. A. Lee P.
Dynamical Evolution of Escaping Lunar Ejecta

As more and more meteorites from the Moon and Mars are discovered, it is
increasingly clear that the ejection of fragments from large bodies is
easier than previously believed. Because few, if any, common source
craters are plausible, at least a half dozen separate launch events have
occurred in the last 10^5-10^6 years (Warrell, 1994, Icarus, in press).
Since the lunar meteorite inventory is undoubtedly incomplete, one can
argue for a reasonably large injected flux of lunar material into the
near-Earth environs.

We have undertaken a detailed study of the dynamical evolution of this
ejected material through the use of numerical simulations. We bombard
the Moon randomly with impactors that eject material at a variety of
velocities above the lunar escape velocity. We find that some fraction
of this material is delivered directly to the Earth, but most escapes
quickly onto heliocentric orbits (as found by Gault, 1983, LPSC XIV,
abstract). The material that reaches heliocentric orbit is then followed
using the regularized mixed variable symplectic integration package of
Levison and Duncan (1994, Icarus, 108, 18-36). We produce statistics
regarding the fraction of the material eventually accreted by the Earth
(and other planets), including the timescales involved, the entry
velocity, and geometry. We hope to ultimately build a self-consistent
model taking into account the primary impactor flux onto the Moon,
launch efficiency for escaping material, and the available data from the
lunar meteorites (cosmic ray exposure ages, strengths, sampling biases,

12.04 Stern S. A.* Flynn B. C.
High-Resolution Imaging of the Lunar Sodium Exosphere I:
Observing Technique

The lunar Na atmosphere has been observed from Earth since its discovery
in 1988. Both spectroscopy and wide-angle imaging has been used. We
present a new imaging technique that takes advantage of a narrow-band Na
interference filter, unique observing geometry, and good internal
scattered light suppression to enable probing of the atmosphere over the
lunar terminator. In this paper we describe our imaging technique and
present clear evidence of this method's success at detecting the lunar
Na exosphere. In paper II we present results of analysis of our data.

The narrow-field (4x4 arcmin) imaging technique involves taking images
just to the dark side of the lunar terminator with a 1-m-class telescope
and a CCD. The lunar surface within the field-of-view (FOV) is then
illuminated only by earthshine, allowing sufficient contrast for
detecting atmospheric Na. By taking image strips along the terminator as
the terminator moves across the lunar surface, the structure of the
atmosphere can be mapped out over approximately two thirds of the lunar
surface visible from the Earth. By covering both lunar mare and highland
surface units, spatial variations due to inhomogeneous surface
composition may be characterized. Furthermore, temporal variations due
to dynamical transport across the FOV can be detected by taking a
time-sequence of images over the same surface regions.

The imaging technique described here complements other methods of
observing the lunar atmosphere. Whereas spectroscopy is limited to
observations off the lunar limb, the narrow-field technique can measure
the atmosphere over roughly 70% of the Earth-facing hemisphere.
Furthermore, wide-angle coronagraphic imaging has so far observed only
the extended lunar atmosphere at distances greater than 0.5 lunar radii
from the surface, whereas our narrow-field imaging technique can observe
the atmosphere essentially down to the surface.

12.05 Simonelli D. P.* Veverka J. Thomas P. C. Helfenstein P.
Switala A. Wisz M.
Photometric Variations on Phobos: Reexamining the Viking Images

Using techniques developed for analysis of Galileo Gaspra and Ida data
(Simonelli et al. 1993, Icarus 103, 49-61) we have corrected
clear-filter Viking images of Phobos for global-average limb-darkening
and phase-angle effects, to investigate the range of albedos and extent
of photometric diversity on the satellite. Normal reflectances range
between ~ 0.06 and 0.10, with a mean value just above 0.07. Most of the
surface has a normal reflectance near the global mean value; the higher
reflectances are associated with the small fraction of the surface
comprised of bright crater rims. Photometrically corrected images
covering phase angles 1.5 degrees -47 degrees reveal that the contrast
between the bright northeast rim of Stickney and the crater floor
increases significantly with rising phase, a trend opposite to that seen
for bright rims around small craters. Preliminary indications are that
the Stickney contrast changes are due more to an atypical photometric
behavior of the crater floor than to any unusual scattering behavior of
Stickney's rim; Stickney's floor may have particles slightly more
backscattering than average Phobos, or may have a higher than average
macroscopic roughness. Because slumping has considerably modified
Stickney' s floor, the processes that roughen this surface might be
different from those proposed for dark markings in fresh crater floors
elsewhere on the satellite (Goguen et al. 1978, GRL 5, 981-984). The
photometric corrections described above were performed with a
global-average Hapke function determined by fitting disk-resolved data
from 22 clear-filter images (phase angles 1.5 degrees to 123 degrees).
The average macroscopic roughness is well-constrained at 22 degrees,
making Phobos somewhat smoother than Gaspra and comparable in roughness
to the Moon. The results also confirm earlier indications that Phobos
has a strong opposition surge-stronger, in fact, than those of Gaspra or
the Moon.

This work was supported by NASA grant NAGW-2186.

12.06 Hapke B. W.* Nelson R. M. Smythe W. D. Horne L. J.
Herrera P. Gharakanian V.
Preliminary Results of Studies of the Opposition Effect Using
the JPL Long-Arm Photopolarimeter

Using the new JPL goniometric photopolarimeter, which can measure the
orthogonal components of reflectances of materials in linearly and
circularly polarized light at phase angles as small as 0.2 degrees, we
have studied the opposition effect (OE) in lunar soils and terrestrial
powders. Previous measurements of lunar soils at phase angles larger
than 1 degree showed that the lunar OE is caused primarily by coherent

High albedo powders have large OE's that are concave-upwards and
independent of grain size. In low albedo powders the OE peak is
truncated and flattened, as predicted by the coherent backscatter model
of the OE. Major exceptions to the latter are the lunar soil samples,
for which the OE's are concave-upward, with surges as large as 66%. The
amplitudes of these OE peaks decrease as albedo increases, directly
contrary to predictions of the coherent backscatter model; this implies
that part of the lunar OE is caused by shadow hiding.

The typical planetary, well-developed, wide branch of negative
polarization, such as the moon's, is associated with a circular
polarization ratio that decreases as the phase angle decreases, implying
that this negative polarization is caused by shadow hiding.

This research is support by grants from the NASA PG&G program.

12.07 Morgan T. H.* Killen R. M.
Unsteady Diffusive Flows in the Mercurian Regolith

Deep (~10 km) crustal release of gaseous Na and K has been invoked as
both the source for these species in the exospheres of the Moon and
Mercury and to explain observed temporal changes [1,2 and references
therein]. Here we examine the temporal behavior of the flux at the
surface that results from the sudden release of a large quantity of gas
at depth in response to some crustal forcing event. With D(z) = D(sub)0
exp(z/lambda) [2] where z is the distance from the base of the crust and
lambda is e-folding distance, and the loss at the surface of the
regolith = Au(sub)s = -D{Du/Dx}(sub)s where u is the density of Na or K
in the pores, we show that the Crank-Nickelson difference equation for
the density, u(sub)j between any two consecutive time steps can be

Equations 1 and 2 appear here in the hard copy.

We examine solutions of these finite difference equations looking at the
temporal behavior in terms of the quantity (L2/4* bar-D) in order to
examine the possibility that deep crustal release could explain the
observed temporal behavior of K.

Supported by NASA grant NAGW 3824.

References: [1] Sprague A. L. et al. (l990) Science, 249, 1140. [2]
Killen R. M. and Morgan T. H. (1993) Icarus, 101, 293.

12.08 Kozlowski R. W. H.* Sprague A. L. Witteborn F. C.
Cruikshank D. P. Wooden D. H.
Mercury: Mid-Infrared (7.3-13.5 Micrometer) Spectroscopic
Observations Showing Features Characteristic of Plagioclase

Mid-infrared spectroscopic observations of the surface of Mercury are
reported for the wavelength range 7.3 to 13.5 micrometers. The observed
spectral radiance emanated from equatorial and low latitude regions
between 110-130 degrees mercurian longitude. The area is primarily an
intercrater plain. The spectra show distinct and recognizable features,
the principal Chnstiansen emission peak being the most prominent. The
Chnstiansen feature and overall spectral shape strongly suggests the
presence of plagioclase (Ca,Na)(Al,Si)AlSi(sub)2O(sub)8. The emissivity
maximum of albite typically falls near 7.7 and that of anonhite near 8.2
micrometers; the Mercury emission maximum is at 8.0 micrometers.
Assuming a linear mixing model, and making adjustments for the vacuum
environment, this corresponds to Ab(sub)(30-10) (bytownite). Labradonte
is not ruled out. Both labradorite and bytownite are characteristic of
igneous rocks of gabbroic composition and of anorthosites.

Mercury observations were made at the NASA Infrared Telescope Facility
(IRTF) 3.0 m telescope on Mauna Kea, Hawaii. The thermal IR spectra were
obtained July 12, 1992 using the High Efficiency Infrared Faint Object
Grating Spectrometer (HIFOGS) Witteborn et al. [1]. HIFOCS has
cryogenicailly cooled aperture and filter wheels, grating, and a linear
array of 120 Bi-doped Si detectors. To increase the spectral range
slightly, and to increase the resolution (decrease the wavelength
sampling increment), all standard stars and Mercury were measured at two
grating settings. The first grating setting spanning wavelengths of 7.22
to 13.02 micrometers and the second 7.61 to 13.30 micrometers. The
individual spectra have an average resolution of 0.048 micrometers.
After the spectra from the two grating settings are interleaved, the
spectral resolution is improved to 0.025 micrometers. Telluric
absorptions were corrected for by using alpha Boo. All ratioed spectra
were then corrected for the spectral shape of alpha Boo (K2 II) using a
spectrum generated from several observations of alpha Boo from Kuiper
Airbome Observatory, the IRTF, and the NASA 1.5 m telescope at Mt.
Lemmon (Cohen et al. l 2]; Cohen et al.[3]).

References: [1] Witteborn F.C., et al. Astro. Soc. Pacific Conf. Series
Astro. IR Spec. Conf. 365-372 (1993). [2] Cohen et al. Astro. J. 104, 5,
2030-2044 (1992). [3] Cohen et al. Astro. J. 104, 5, 2045-2052 (1992).

12.09 Butler B. J.* Muhleman D. O. Slade M. A.
VLA/Goldstone 3.5-cm Radar Observations of Mercury in 1994:
South Polar and Other Results

The discovery of deposits of ices at the poles of Mercury has profound
implications for planetary science [1-4]. Studies of these deposits may
provide clues about the distribution and type of impactors and the
supply of volatiles in the early inner solar system, the state of the
spin vector of Mercury in the past, and the thermal history of Mercury,
among other things. The deposits have been mapped at both poles by the
Arecibo radar telescope at 13-cm [2,4]. The north polar deposits have
also been mapped at 3.5-cm [1,3]. In an attempt to map and characterize
the south polar deposits at 3.5-cm, we performed two experiments when
the south polar regions were visible. These experiments took place on
Feb. 21 and 26 of 1994, when the subearth latitude was ~-10 degrees,
giving a good view down into regions on the surface which remain in
permanent shadow, the most likely regions to find the polar ice deposits
[3]. Because of the extreme complexity of the data reduction involved in
these experiments, the data have not been completely reduced yet.
However, preliminary results indicate the presence of a south polar
feature similar to that at the north pole. The peak radar cross section
received in the same circular polarization as that transmitted (SS) on
Feb. 21 was ~ 6.6%. Also seen in the SS data for that day are large
cross section enhancements which we have called "radar basins" [3], one
of which is located near the Kuiper crater. These features and their
cause will be discussed, along with the south polar feature.

1 - Slade, Butler, and Muhleman, Science, 258, 635, 1992. 2 - Harmon and
Slade, Science, 258, 640, 1992. 3- Butler, Muhleman, and Slade, JGR, 98,
15003, 1993. 4 - Harmon, et al., Nature, 36g, 213, 1994

12.10-P Magnusson P. Lumme K.
Photomorphography Applied to Rough and Globally Concave Objects

Photomorphography is a technique for inverslon of dlsk-integrated
photometrlc observations to derlve global characterlstics of solid Solar
System bodies (see Kaasalainen, Lamberg, Lumme, and Bowell 1992, Astron.
Astrophys. 259, 318-332). A basic assumptlon ln thls technlque is that
the object has a convex shape. Relaxation of this assumption makes the
whole problem mathematically very intractable.

We are studying the sensitivity of our results to departure from the
convexity assumption. For example, for a simple model with homogeneous
scattering and a global concavity on the equator the resulting synthetlc
lightcurves at non-zero solar phases have a significant component in the
first Fourier order. When interpreting this in terms of convex models
one erroneously arrives at a strong albedo contrast and a shape that
resembles neither the concave object nor its convex hull. We wlll
dlscuss ways of overcoming this problem, or at least diagnosticise when
there is a problem.

12.11-P Robertson S. Walch B. Horanyi M.
Charging Characteristics of Simulated Lunar Regolith Dust
Grains in a Plasma

The charge on grains of simulated lunar regolith (MLS-1) has been
measured in a plasma environment containing thermal electrons of a few
eV temperature and a second component of nearly monoenergetic electrons
with energy variable from 15 to 120 eV. Measurements are compared with
the charging characteristics of glass spheres and copper dust in the
same range of size (53-63 microns). The charging characteristics can be
divided into three regimes: one in which the grains charge to a
potential which repels the bulk of the thermal electrons, a regime of
higher potential in which the thermal and the fast electrons are
repelled, and a bi-modal regime in which grains may charge to either
potential. The charging is described by a model which includes the
currents of the thermal electrons, thermal ions, fast electrons and
secondary emission induced by the fast electrons.

12.12-P Potter A. E. Morgan T. H. Gilliam L.
Coronagraphic Observations of the Lunar Sodium Exosphere Near
the Lunar Surface

Imaging of the sodium exosphere was first accomplished by Flynn and
Mendillo (Science, 261, 184 (1933)) using optical filters to isolate the
sodium emission and occulting masks to block scattered moonlight. Their
technique was limited to regions above about 1000 km., because of
interference by scattered moonlight in the optics and sky. Images of the
sodium exosphere near the surface are needed to resolve issues
surrounding the origin of the sodium and its interaction with the
surface (see Potter and Morgan, GRL, 1994, in press). In order to image
the sodium exosphere near the lunar surface using the optical filter
technique, extremely high rejection factors for scattered light are
required. These are found only in special purpose instruments like the
Evans 40 cm.,solar coronagraph at the National Solar Observatory at
Sacramento Peak New Mexico. We report initial observations of the lunar
sodium exosphere near the lunar surface, using the Evans coronagraph, a
2K x 2K CCD, and a 3 A bandwidth interference filter. Our results
illustrate the difficulties involved in imaging observations close to
the lunar surface, as well as the importance of these data to our
understanding of the processes which form and modify the lunar sodium

The work was supported by the NASA Planetary Astronomy program.

12.13-P Belobrov A. V. Fuks I. M.
Approximation of the Scattering Data from the Moon

Recent theoretical results in the problem of the electromagnetic wave
scattering by a rough surface offer new possibilities to determine
interface parameters from the experimental data. Statistical parameters
of the Moon's surface, e.g. spacial spectrum of the irregularities, has
been found from comparison between the experimental and computational
data obtained in the framework of different surface models. These
results were compared with the well-known Hagfor's approximation model

12.14-P Nelson R. M. Horn L. J. Weiss J. R. Smythe W. D. Wright F.
Hermes-The Mercury Orbiter Discovery Mission

The Hermes Mercury orbiter is one of several missions that have been
proposed to NASA under the Discovery program. If selected, Hermes would
be launched in 1999 and would be inserted in a 200X15000 km., 12hr,
elliptical orbit about Mercury. The tour to Mercury would include two
Venus-Mercury flybys.

The principal objectives of the mission are: surface mapping, surface
composition and texture determination, mapping the magnetic and
gravitational fields, understanding the distribution and abundances of
atmospheric species, characterizing the species in Mercury's
magnetosphere, and understanding the planet's interior.

There are four instruments under consideration for the Hermes payload.
These are: an imaging system, a magnetometer, an ultraviolet
spectrometer, and a lidar. The imaging system will provide the
measurements intended to study the large scale surface morphology and to
distinguish between impact and volcanic landforms. Along with the
ultraviolet spectrometer, it will be able to provide information on
questions regarding the abundance of iron in the planet's crust. Both
systems are able to search for water ice at Mercury's poles. The camera
and the lidar provide several ways to measure the scattering properties
of the regolith, and hence infer its texture. The Ultraviolet
spectrometer will measure the abundance of species in the planet's
tenuous atmosphere. The particle and fields investigations will provide
information about the nature of the planet's interior and will
characterize the material in the planet's magnetosphere. This work
carried out at JPL under contract with NASA.

12.15-P Lopes-Gautier R. Baloga S. Nelson R.
Volcanism on Mercury

No volcanic features have yet been unambiguously identified on Mercury.
However, Mercury presents an unusual opportunity to investigate the role
of planetary volcanism in comparison with other planetary surface
processes (e.g. cratering, thermal and orbital evolution) and
constraints (e.g. composition, shape, and interior configuration).
Advances in the understanding of effusive volcanism over the last decade
now make it possible to constrain eruption rates, compositions,
rheologic properties, thermal characteristics, and eruption style from
dimensional and topographic information.

We will elaborate these quantitative aspects of volcanism on Mercury on
regional and global basis in light of the recent improvements in the
current theories of lava emplacement. Particular attention is paid to
the influence of regional differences in relation to compositional and
topographic influences and the competition of volcanism with other
surface processes. These results provide additional constraints on the
orbital and thermal evolution of the planet and the interior processes.
Presently, the ability to understand the interrelationship of planetary
processes for Mercury is lirnited by data, rather than the theoretical
capabilities. We will discuss potential remote sensing measurements that
would improve the understanding of volcanism in relation to the other
processes that formed the surface and interior processes. Presently, the
ability to understand the interrelationship of planetary processes for
Mercury is limited by the data, rather than the theoretical
capabilities. We will discuss potential remote sensing measurements that
would improve the understanding of volcanism in relation to the other
processes that formed the surface and interior of the planet.

12.16-P Flynn B. C. Stern S. A.
High-Resolution Imaging of the Lunar Sodium Exosphere II:
Science Results

We present results of analysis of lunar atmospheric imaging data
obtained with the University of Texas McDonald Observatory 0.9-m
telescope equipped with a CCD. Our imaging technique, described in paper
I, employs high-resolution, narrow bandpass, and specific observing
geometry to suppress scattered light and image the lunar atmospheric Na
essentially down to the surface.

We have conducted observations at McDonald with this technique in 1991,
1993, and 1994. We have analyzed the data by applying a simple
collisionless exosphere model to obtain temperature and spatial
distribution information. The model assumes spherical symmetry and a
Maxwellian distribution. Image analysis completed to date yields
evidence of (1) coexistent cold (thermal) and hot (suprathermal) Na
populations; and (2) systematic variation in atmospheric temperature and
surface fractional abundances with latitude.

The data taken to date represent <10% of the visible lunar surface.
Future observations will concentrate on obtaining data over the full
extent of lunar surface that can be observed during the first
quarter-full Moon and full Moon-last quarter observing windows. We will
use our imaging technique to (1) make lunar Na "movies" over specific
surface areas; (2) make full north-south Na image strips along the
terminator over a wide range of lunar phases; and (3) extend imaging
observations to lunar K, which has thus far only been detected through

Tuesday, 1:30 - 2:15 Crystal Ballroom
M. A. McGrath, Moderator

12A.01-INV Gladstone G. R.*
Planetary Science in the Extreme Ultraviolet

No abstract available.

Tuesday, 2:15 - 3:00 Crystal Ballroom
D. M. Hunten, Moderator

12B.01-INV Ingersoll A. P.*
S-L 9: What Happened?

No abstract available.

SESSION 13 ....... Jupiter II and Saturn
Tuesday, 3:30 - 5:30 Crystal Ballroom A
C. D. Barnet and B. Sicardy, Moderators

13.01 Bezard B.* Griffith C. Owen T. Lacy J.
Search for Hydrogen Cyanide on Jupiter

We present high spectral resolution observations of Jupiter recorded on
January 21, 1991 at the NASA/Infrared Telescope Facility . The cryogenic
echelle array spectrometer IRSHELL was used at a resolution of 0.08
cm^-l to search for the R(5) and R(ll) lines of HCN at 735.61 and 747.41
cm^-l. Signal-to-noise ratios of several hundred were achieved in the
continuum. We do not find evidence for HCN absorption, in contrast to
the pioneer observations of Tokunaga et al. (1981, Icarus 48, 283-289).
Synthetic spectra have been generated using a HCN vertical profile
limited by saturation in the upper troposphere. The calculated lines
yield a poor agreement with Tokunaga et al.'s data for any value of the
deep HCN mixing ratio. From our non-detection of the HCN lines and our
re-analysis of Tokunaga et al.'s observations, we think that the
previously reported detection of HCN in Jupiter is questionable.

13.02 Weisstein E. W.* Serabyn E.
Detection of the PH3 J=3-2 Rotational Transition in Jupiter and

We report the detection of absorption in the J = 3-2 lines of PH(sub)3
at 800.5 GHz in both Jupiter and Saturn. While PH(sub)3 has been
detected in the infrared in both Jupiter and Saturn, and at mm
wavelengths in Saturn (Weisstein and Serabyn 1994), the measurements
presented here represent the first submillimeter confirmation of the
presence of PH3 in Jupiter.

Jupiter and Saturn were observed at the CSO in June 1994 with our FTS. A
strong ~20% absorption feature with a FWHM ~ 10 GHz and line center of
800.6 GHz was observed in Jupiter. An even broader feature was detected
at the same frequency in Saturn, but its large linewidth, combined with
the cutoff in terrestrial atmospheric transition near 780 GHz, make this
line's parameters more difflcult to determine.

The observed lines are modeled in both giant planets using a radiative
transfer code, and results concerning the distribution of PH(sub)3 in
the Saturnian and Jovian stratospheres and upper tropospheres are
derived. In the case of Saturn, additional constraints are provided by
the previous detection of the J = 1-0 line. The enrichment of PH(sub)3
in Saturn relative to Jupiter is confirmed.

In addition, new low-noise FTS continuum observations of Jupiter and
Saturn between 330 and 950 GHz have yielded no further line detections,
allowing new upper limits to be placed on the mixing ratios of species
such as CO, HCN, HCP, and H(sub)2S in their upper atmospheres.

13.03 Varanasi P.* Li Z. Nemtchinov V.
Laboratory Data on the IR Lines of 12CH4, 13CH4, 12CH3D,
12C2H2, and 12C13CH2 Observable in the Jovian Atmosphere

Infrared observations of the Jovian atmosphere involve the study of the
infrared spectral emission by CH(sub)4, its isotopes, and hydro-carbons
which are created by its photodissociation. As laboratory data are
needed on the spectral lines of interest at the appropriate temperatures
to aid the astronomers, we present absolute intensities,
collision-broadened half-widths, and pressure-induced shifts on lines
belonging to thermal infrared bands of ^l2CH(sub)4, ^l3CH(sub)4,
^l2CH(sub)3D, ^l2C(sub)2H(sub)2 and ^l2C^l3CH(sub)2 obtained with a
Fourier-transform spectrometer. Using He and H(sub)2 as broadening
gases, the data have been measured at temperatures appropriate to the
planetary atmosphere.

13.04 Romani P. N.* Monks P. S. Nesbitt F. L. Payne W. Stief L. J.
Fahr A. Laufer A. H.
Recent Rate and Product Measurements of the Reactions C2H3 + H2
and C2H3 + H--Importance for Photochemical Modeling of
Hydrocarbons on the Outer Planets

Recently the rate constant for the reaction C2H3 + H2 has been measured
by a more direct method than any used previously. While only an upper
limit at room temperature was determined, the rate is lower by at least
a factor of 400 than that used in Allen et al. 1992 (Icarus, 100, 527).
This reaction is potentially important in regulating the relative
abundances of C2H2, C2H4, and C2H6 in the stratospheres of the Jovian
planets. Specifically, this reaction converts C2H3 to C2H4 before it is
either recycled to C2H2 or reacts further to form C3 and C4
hydrocarbons. The resultant C2H4 can be transformed to C2H6 by
successive reactions with H or photodissociated back to C2H2. While the
new rate constant for C2H3 + H is not significantly different from
previous measurements, the determination of the yield of the combination
product C2H4 (~25% at room temperature and ~1 mbar) is significant.
Previously, the products were assumed to be solely C2H2 + H2, i.e.,
recycling C2H3 before it could be converted to C2H4. For photochemical
modeling, the reduced conversion rate of C2H3 to C2H4 via H2 is
partially offset by inclusion of the C2H4 product channel for C2H3 + H.
We will present results from a 1-D hydrocarbon photochemical model with
an emphasis on sensitivity to the extrapolation method chosen to bridge
the pressure/temperature regime of laboratory measurements to those in
the stratospheres of the outer planets.

13.05 Drossart P.* Forni O. Beuzit J. L. Baines K. Orton G.
Structure of the Polar Haze of Jupiter from Ground-based
Observations in 1994

Observations of the South polar regions of Jupiter at 2 microns have
been conducted on 1994 March 27 and 29 with the adaptive optics system
COME-ON+ on the 3.60 meter ESO telescope. By using a galilean satellite
as a wavefront reference, as well as for accurate offset guiding, a
spatial resolution better than 0.2 arcsec (for the PSF width) was
obtained on Jupiter images at 2.07 and 2.12 micron, in the dipole
induced absorption band of H(sub)2. At these wavelengths, the upper
atmospheric levels of Jupiter are sounded between 5 and 70 mbar (Kim et
al, Icarus, 1991). At this spatial resolution, the South polar haze
appears to be non-uniform, and spatial structures in the polar haze are
observed, corotating with Jupiter .

A comparison will be made with images obtained one month later with
NSFCAM on IRTF. These images are obtained during the nights of April 25
and 28. A global longitudinal survey has been obtained, at selected
wavelengths between 1 and 5 micron. Morphological studies of the polar
haze, in particular, can be made from these images, both in North and
South polar regions. From the observations of each night, cylindrical
mosaics can be built and cloud tracking will allow us to study the
differential rotation of Jupiter at several wavelengths in the infrared.

13.06 Barnet C. D.* Caldwell J. J. Cunningham C. C.
HST Spectra of Jupiter and Saturn: Characteristics of
Stratospheric and Tropospheric Hazes

Observations of Jupiter (May 1992-July 1993) and Saturn (December 1992)
were made by the HST Faint Object Spectrograph (FOS) (1800-3300
angstrom) and the Goddard High Resolution Spectrograph (GHRS) (1600-1850
angstrom). The planetary spectra sample the equatorial central meridian,
equatorial limb, and (Saturn only) the north pole.

The FOS spectra were processed to remove background grating scatter and
to account properly for effects due to spherical aberration in the HST
primary optics. The planetary spectra were ratioed to HST spectra of the
solar analog 16 Cyg B over the spectral range 1600-3300 angstrom. HST
images at 3360 angstrom are used to derive limb-to-limb absolute
reflectivity scans, I/F, for both Jupiter and Saturn.

The absolute reflectivity ratios are compared to vertically
inhomoge-neous models of the planetary atmospheres to determine
abundances of minor and trace species therein. Both Jupiter and Saturn
show the influ-ence of acetylene and phosphine absorption below 2300
angstrom. Only Jupiter shows features due to ammonia. Preliminary models
of these data were discussed by Caldwell et al. (1993, BAAS 25, 1027).

In this paper we will concentrate on the vertical distribution and
op-tical characteristics of the hazes in these atmospheres. The
wavelength dependent models at FOS wavelengths are most sensitive to
aerosol parameters (e.g., particle size distribution, refractive index)
and gas concentrations. The geometry dependent models of limb-to-limb
reflec-tivity characteristics are most sensitive to particulate phase
functions. Simultaneous utilization of both types of models are used to
optimize our understanding of these atmospheres.

13.07 Gudkova T.* Mosser B. Gautier D. Guillot T. Provost J.
Chabrier G.
Jovian Seismology: Influence of the Troposphere Thermal
Signature and Seismological Diagnosis

In order to correctly interpret the seismological measurements recor\ded
after the cometary impact on Jupiter, we have examined the following
1) The Jovian oscillation spectrum taking into account the planetary
troposphere and stratosphere has been calculated for the first time.
Direct numerical computations of planetary eigenfrequency pattern have
been carried out for the fundamental modes and for non radial modes with
radial order up to 20 and degree up to 30. The differences of the
oscillation frequencies due to the addition of the tropospheric and the
stratospheric levels strongly depend on the upper boundary condition.
However, the corrections are simple functions of the eigenfrequency and
the degree of the mode. These results are necessary for accurately
interpreting the future seismological observations of Jupiter and

2) The propagation of acoustic waves in the Jovian troposphere has been
revisited, in order to estimate their thermal signature. The relation
between this signature and the wave velocity depends strongly on the
wave frequency and on the level of detection.

3) Different interior structure.models of Jupiter have been compared
with the help of seismology. We have considered two groups of models:
first, models based on different equation of state of hydrogen, and
second, models with different descriptions of the mechanism of energy
transport (fully adiabatic models and models with a radiative window).
Our results show it is possible to test the giant planets interiors with
the help of future seismological data, depending on the accuracy of the
determination of the eigenfrequencies and on the degrees of the observed

13.08 Hubbard W. B.* Porco C. C. Hunten D. M. Rieke G. H.
Rieke M. J. McCarthy D. W. Haemmerle V. Haller J. McLeod B.
Lebofsky L. A. Marcialis R. Holberg J. B. Landau R.
Carrasco L. Elias J. Buie M. W. Persson S. E. Boroson T.
West S. French R. G.
Saturn's Mesospheric Temperature from 28 Sgr Occultations

A full ensemble of immersion/emersion lightcurves from the occultation
of 28 Sgr by Saturn in 1989 has been analyzed. From an accurate
determination of the Saturn center of figure using fiducial ring
features, the center is known a priori to within a few km. Thus we
retrieve not only atmospheric temperatures but also absolute radial
positions of the measurements with high accuracy. From the 11
best-calibrated lightcurves, we infer a mean temperature of 135 +- 5 K
on a level surface corresponding to an equatorial radius of 60963 +- 3
km. The pressure corresponding to this level (isothermal fit) is 1.7
microbar. Saturn-centered latitudes ranging from -14.20 degrees to +6.25
degrees were probed. To fit the actual shape of the planet, we assumed
that it was rotating on cylinders with the observed differential flow
field as derived from cloud motions. There is considerable "topography"
(several x 10 km) in the level surfaces over the range of latitudes
probed, and this effect is included in the model. The self-consistencies
of the inferred level-surface radii at half flux for each of the
stations provide support.for the differentially-rotating model. Two of
the stations, CTIO and Las Campanas, observed immersion through C-ring
features. The features thus acted as a fiducial grating superimposed
over the atmospheric modulation at these stations. Model lightcurves
computed on the basis of the adopted mean temperature and
center-of-figure solution agree very well with the data from these

This work has been supported by NASA grants NAGW-1555, NAGW-1876,
NAGW-1368, by NSF Grant AST-8715373, and by others.

13.09 Allison M.*
A Lower Isentropic, Iso-Potential Vorticity Model for the Deep
Saturn Thermocline

A simple conceptual model for the zonal winds of the outer planets is
proposed with particular attention to the deep thermal structure of
Saturn. A dynamic upper layer of well-mixed potential vorticity is
assumed, with a static stability exponentially decreasing with height
above a windless, isentropic interior. The schematic character of the
geostrophically balanced potential temperature field is shown by the
solid contours in the figure, with isotherms dashed (for every 20K).
Quasi-uniform potential vorticity for the configured thermocline is
consistent with a high-latitude alternation of jets over the interal
deformation scale and cyclonic shear toward a strong equatorial jet.

13.10 Bosh A. S.*
Saturn's Gravitational Harmonics

We present a solution for Saturn's pole and radius scale derived from S
occultation data sets, including 2 recent data sets acquired using the
High Speed Photometer on the Hubble Space Telescope. This solution is
used in an investigation of kinematic models for non-circular ringlets.
Constraints from the location of the Titan ringlet (in an apsidal
resonance with Titan) and the measured precession rates of the Maxwell
and 1.495 Rs ringlets are combined with previously published constraints
from Pioneer tracking data and satellite precession rate measurements
(Null et al., 1981. Astron. J. 86, 456-468) to determine values for
Saturn's gravitational harmonics: J(sub)2 = (16301 +/- 6) x lO^-6,
J(sub)4 = (-894 +/- 9) x lO^-6, and J(sub) = (124 +/- 5) x 10^-6 (for
Saturn equatorial radius 60330 km). Values of higher-order harmonics are
held fixed: J(sub)8 = -10 X l0^-6, J(sub)10 = 2 X 10 ^-6, J(sub)12 =
-0.5 X 10^-6, J(sub)14 = J(sub)16 =...= 0. The formal errors of the
first three even gravitational harmonics are greatly reduced from those
of Nicholson and Porco (1988, J. Geophys. Res. 93, 10209-10224),
particularly the error in J(sub)6. J6 is now determined precisely enough
to prDvide a useful constraint on models of Saturn's interior. The value
we find for this quantity suggests that the interior may be in a state
of complex rotation.

13.11-P Martyn M. D. Barnet C. D. Caldwell J. J. Hopkinson J. M.
HST Spectra and Image of Saturn. Interpretation

A Hubble Space Telescope (HSI) ultraviolet (F336W filter) images of
Saturn in December 1992 shows a brightening on the limb at 35 degrees
north latitude. At the time of these observations the phase angle was
4.85 degrees and the limb was on the sunset (east in system III) side of
the planet. Reflectivity scans show that the limb is 4% brighter than
the central meridian (cm) and the terminator is 15% brighter than the

We have utilized a vertically inhomogeneous model of Saturn's
atmo-sphere to synthesize the absolute reflectivity, I/F. The model was
opti-mized to fit HST/FOS spectra in the wavelength range of 1800-3200
angstrom for the equatorial and polar regions during the same
observations (Caldwell et al. 1993, BAAS 25,1028). Limb-to-limb scans
from the F336W image constrain the phase function of the hazes used in
the model, while the FOS spectra constrain the altitude, optical depth,
and single scattering albedo of the hazes. We have modeled limb-to-limb
scans and FOS spectra taken at the same latitudes and extrapolated these
models to show that the increase in reflectivity at 35 degrees latitude
is consistent with the observational geometry.

The imaging observations also, confirm differences in vertical structure
between the north polar region and the equatorial regions.

M. D. Martyn was supported by Natural Science and Engineering Research
Council (NSERC) of Canada grant no. 5-50455.

13.12-P Ortiz J. L. Moreno F. Molina A.
Methane Band Imaging of Saturn's Cloud Temporal Changes from
1991 to 1993

We present the latitudinal behavior of absolute reflectivity, molecular
absorption and limb-darkening as derived from CCD Saturn images obtained
at 6190 Angstrom, 6350 Angstrom, 7250 Angstrom, 7500 Angstrom, 8900
Angstrom, and 9500 Angstrom during four observing runs during the period
1991-1993. The observations were made at the 1.52-m telescope at Calar
Alto and the 2.5-m Nordic Optical Telescope at La Palma. The
limb-to-limb reflectivity scans at planetographic latitudes 0 degrees ,
10 degrees , 20 degrees, 40 degrees, and 70 degrees have been analyzed
in terms of multiple scattering radiative transfer models. The large
changes in reflectivity at equatorial latitudes from 1991 to 1992 are
interpreted as a considerable increase in the aerosol specific abundance
at different levels within the extended tropospheric haze.

13.13-P Hillman J. J. Glenar D. A. Le Louarn M. Drummond J. D.
Multispectral Imaging of Saturn Using the Kirtland 1.5 m
Adaptive Optics Telescope

Recent high spectral and spatial resolution imaging results obtained
from observations of Saturn are shown and discussed in a preliminary
fashion. These observational data were acquired using a newly developed
multispectral array camera. Wavelength tuning is achieved by an
acousto-optic tunable filter operating in the 0.6 to 1.0 micron region.
Images were recently obtained using the 1.5 m laser beacon adaptive
optics facility at the Starfire Optical Range, Albuquerque, NM, using a
new image correlator intended for tracking extended (solar system)
sources. Spectral images were taken in H2 (800 nm), NH3 (770 nm), and
CH4 (725 and 890 nm), and corresponding adjacent continuum regions.
Preliminary analysis of this data and a comparison with earlier works
will be made.

This work was supported by the NASA Planetary Instrument Definition and
Development Program and the GSFC Director's Discretionary Fund.

13.14-P Momary T. W. Baines K. H.
Saturn's Tropospheric and Stratospheric Aerosols: Probing
Vertical Structure of Various Latitude Regions with an Analysis
of Near-Infrared IRTF ProtoCAM Imagery

Near-infrared imagery of Saturn acquired in August 1993 in H2 and CH4
absorption bands between 1.60 - 3.90 micrometers affords a means of
probing the vertical distributions of aerosol layers in Saturn's
troposphere and stratosphere. Our objective is to determine vertical
structure and near-infrared optical properties of Saturn's aerosols,
such as single scattering albedos and opacities, as well as to derive
particle sizes and mass column abundances. We have initially used a
simplified three-layer atmospheric model containing a clear gas layer,
an NH3 haze layer, and an optically-thick, physically-thin NH3 cloud
just below the haze. We have assumed a double Henyey-Greenstein phase
function for the haze and cloud particles with f = 0.75, g(sub)1 = 0.7,
and g(sub)2 = -0.3, and a H2 unsaturated molar fraction of 0.924. Three
regions have been analyzed (the Equatorial Zone, the North Temperate
region, and the North Tropical region) and preliminary results suggest
that the aerosol burden in the EZ is either much greater and/or extends
to a much higher altitude than in the other two regions. For example,
for a haze opacity of 2.00 (i.e., an optically-significant layer based
on recent findings of Karkoschka and Tomasko, 1992), we find that the EZ
has a hazetop near 0.083 bars while the NTe and NTr regions are
characterizedby a lower-lying hazetop near 0.160 bars. On the other
hand, if we assume that all regions have a hazetop near 0.160 bars, then
we find that the EZ's opacity is about 2.5 times that of the other
regions. Tentatively, we find that the haze opacity for the NTe and NTr
regions is less than 4.00, from which we can estimate that the particle
radii are significantly smaller than 2 micrometers, consistent with the
conclusions of Karkoschka and Tomasko (Icarus 106, 428-441, 1993). We
are continuing to analyze these images, incolpoIating other wavelengths
(such as an intermediatewavelength on the shoulder of the H2 1-0 S(l)
pressure-induced absorption feature) and enhancing the model, in order
to establish haze structures of these various axisymmetric regions on
Saturn more fully.

13.15-P Del Rizzo D. A. Barnet C. D. Caldwell J. J.
Saturn's Northern Polar Hexagon: Image Analysis and Color

The Voyager spacecraft encounters with Saturn both occurred near the
northern spring equinox and, therefore, were not at optimum geometry for
observing Saturn's poles. Even so, Godfrey (1988, Icarus 76, 335) was
able to construct mosaics of images to reveal Saturn's north polar
hexagon and the associated spot at 76 degrees latitude.

We have surveyed the Voyager 2 database and examined previously
un-published images at high sub-spacecraft latitudes (+ 10 degrees to 20
degrees) and high spatial resolution. We have constructed high
resolution polar projections from selected images in the violet and
green filters. The appearance of the hexagon in these filters is
significantly different; we interpret this as being due to a different
vertical distribution of hazes between the region to the south and to
the north of the hexagon. These images are consistent with the
interpretation that the hexagon is an atmospheric cloud pattern
associated with a complex planetary wave phenomena (Allison et al.,
1990, Science 247, 1061).

HST observations of the hexagon during Saturn's summer season clearly
show the polar hexagon with its associated spot (Caldwell et al., 1993,
Science 260, 326). Comparison of HST images with Voyager images taken
through similar filters shows that the regions delineated by the hexagon
boundary as well as the polar spot have the same color. This suggests
that the hexagon and the polar spot remained unchanged during the 12
years between the Voyager and HST images, and supports the theory that
the polar spot is linked to the rotation rate of the interior.

We acknowledge the National Space Science Data Center (NSSDC) and the
imaging team leader from the Voyager project, B. A. Smith, for
provid-ing the Voyager images used in this analysis. We also acknowledge
Prof. J.A. Westphal who was the original PI for the HST images
considered here. D.A. Del Rizzo was supported by Natural Science and
Engineering Research Council (NSERC) of Canada grant no. 5-50455.

13.16-P Achterberg R. K. Flasar F. M.
Modeling of Planetary-Scale Waves in Saturn's Atmosphere

Analysis of Voyager IRIS data has revealed the presence of
quasi-stationary planetary scale waves of zonal wavenumber 2 and 4 at
the tropopause level on Saturn (BAAS 25, 1055). These waves are
strongest at latitudes between 30 degrees and 40 degrees north,
decaying rapidly to the north and more slowly to the south, and are seen
in data at 130 mbar pressure level, but not at 270 mbar.

We have used a WKB raytracing model of planetary-scale Rossby wave
propagation on a sphere to understand the behavior of the observed
waves. Rossby waves near the tropopause at 35 degrees N are restricted
to phase velocities between about -30 and +30 m/s (as Rossby waves
propagate westward relative to the mean zonal flow with a maximum
velocity dependent upon the static stability). Meridional confinement of
the waves is caused by variations of the mean zonal flow; the strong
equatorial jet creates a turning point near 30 degrees at which the wave
is reflected. Propagation of the wave north of roughly 40 degrees is
restricted by either a turning point or a critical layer (at which the
wave is stationary with respect to the mean flow and is absorbed)
depending upon its phase velocity. The observed rapid decay to the north
is suggestive of a critical layer. Vertical propagation of upper
trophospheric Rossby waves to pressures below about 200 to 300 mbar is
inhibited by the vertical thermal structure of the atmosphere,
consistent with the observations. The waves are, however, free to
propagate upwards into the stratosphere.

This work was supported by a grant from the NASA Planetary Atmospheres

13.17-P Weir A. L.
A Search for Resonant Wavemodes in Outer Planet Atmospheres:
Preliminary Results

Longitudinally averaged 2D models of the upper troposphere and lower
stratosphere of several outer planet atmospheres have been produced
using temperature and velocity fields obtained during the Voyager
encounters. These are being used to look for resonant equatorial and
midlatitude planetary wavemodes. Results of this search will be
presented and, where relevant, compared to observations.

13.18-P Schinder P. J. Flasar F. M.
Sensitivity of the Radio Occultation Technique to Spacecraft
Transmitter Power

We investigate the sensitivity of the radio occultation technique to the
power of signal produced by the on-board transmitter on the occulted
spacecraft. This is done in anticipation of the Cassini mission to
Saturn, where the S-band transmitter will be operating at roughly half
the power of the Voyager 2 transmitter. We take an atmosphere of known
structure and perform a synthetic occultation of a spacecraft on a known
trajectory emitting a signal of known frequency, constructing a time
series of the complex antenna voltage received by a receiver on the
earth. For convenience, the atmosphere we choose is taken from our
determination of the structure of the atmosphere of Saturn using the
Voyager 2 ingress occultation, and the trajectory that of Voyager 2
during this occultation. The loss of power due to refractive defocusing
of the signal from the spacecraft is fully accounted for as the ray from
the spacecraft passes through the atmosphere. We perform a series of
occultations by adding white noise at the antenna to the pure
monochromatic signal from the space craft. In addition to the no noise
case which forms the baseline, we perform a Monte Carlo series of
occultations at the Voyager 2 free-space signal-to-noise ratio (SNR),
one half that value (roughly the Cassini value), and one quarter that
value (half the Cassini value). After the synthetic time series is
produced, we use our standard radio occultation software to determine
the run of temperature T vs. pressure P of the atmosphere from the
synthetic time series. The end result is a series of plots of Delta T,
the difference in temperature from a particular run from the known T of
the atmosphere vs. P. From this we can determine the degradation in the
retrieved temperature profile with decreasing SNR. Preliminary results
done using the Titan atmosphere show that while the current planned band
transmitter power for Cassini is probably adequate, a lowering of the
power by a factor of two would result in unacceptably large errors.

SESSION 14 ....... Mars Surface
Tuesday, 3:30 - 5:30 Crystal Ballroom B
G. B. Hansen and E. Merenyi, Moderators

14.01 Houben H.* Haberle R. M. Young R. E. Zent A.
The Regolith-Dominated Martian Water Cycle

Martian water vapor sublimes from the permanent north polar cap during the
summer and is transported through the atmosphere to lower latitudes. It is
an open question whether that water returns to the north pole at other
seasons (i.e., whether the current annual cycle is closed) or if it is
being transfered to other reservoirs, such as the south polar cap or the
regolith. To answer this question we have developed a Mars Climate Model-a
simplified atmospheric general circulation model capable of multi-annual
simulations, coupled to a subsurface regolith. We have used this model to
perform a suite of ten-year-long simulations of the current Martian water
cycle. Each simulation begins with a dry atmosphere and a limitless supply
of surface ice polewards of ~ 75 degrees N.

In one set of simulations, all super-saturated atmospheric water vapor
precipitates on the surface as frost where it is available to re-evaporate
whenever conditions allow. In these simulations, the water cycle is not
closed. Large quantities of water are deposited in seasonal frost caps
(extending nearly to the equator in winter) and in a permanent cap near the
south pole. After a decade, all latitudes have > 100 micrometers available
as surface ice or in the atmospheric column. This situation bears little
resemblance to Mars as we know it.

In another set of simulations, surface ice and water vapor in the lowest
atmospheric layer (~ 250 m thick) are in daily equlilibrium with a 1 cm
thick layer of the regolith. In these simulations, adsorption is the
dominant process and no ice deposits form outside the initial source
region. Since the adsorptive capacity of the soil increases with the vapor
pressure, the soil buffers the atmospheric water vapor column. A seasonal
adsorbing layer forms in the northern mid-latitudes into which water is
deposited in summer and from which it returns to the permanent cap in
stages through the remainder of the year. These simulations are
qualitatively quite Mars-like, though the typical water vapor column
abundances of ~ 1 precipitable micrometer are a bit low compared to Viking

14.02 Clifford S. M.*
Mars: The Response of an Ice-Rich Crust to Burial by
Volatile-Poor Materials

Consideration of the martian valley networks, outflow channels, and
various other surface features, suggests that the martian crust is
water-rich and that it may host a planetary inventory of outgassed water
equivalent to a global ocean 0.5 - 1 hn deep [Carr, Nature, 326, 30-35,
1987J. However, the geologic evidence also indicates that Mars has
undergone extensive resurfacing including the production of up to 2 km
of ejecta from impacts, as much as 0.5 km of extrusive volcanics, and an
unknown (but potentially large) volumetric contribution of interbedded
weathering and sedimentary deposits. Therefore, given the influence of
crustal temperature on the stability and distribution of ground ice, and
the effect of deposition on subsurface temperatures, the response of an
initially ice-rich crust to burial by volatile-poor mantles has been
considered - including potential changes in state, distribution, and
modes of volatile transport. At the base of the cryosphere (that region
of the crust where the temperature remains continuously below the
freezing point of water), the added insulation provided by the
emplacement of a depositional layer will result in the rise of the
melting isotherm until thermal equilibrium is reestablished - resulting
in the melting of any intervening ice and the subsequent drainage of
meltwater into the lowermost porous regions of the crust. At the top of
the ice-rich layer, the increase in local temperature will lead to the
movement of the crustal 'cold-trap' into the overlying mantle -- causing
the ice present at depth to be thermally redistributed until it has
saturated the available pore space in the colder regolith, above. This
redistribution may occur by three different processes that involve all
three phases of water, these include: thermal vapor diffusion (where
vapor migrates from the warmer depths to the colder near-surface crust),
thermal liquid transport (in response to the temperature-induced
gradient in soil water potential that can occur in the interfacial films
between rock and ice), and regelation (the movement of ice through soil
pores via pressure induced melting and refreezing). By these processes,
an initially volatile-poor depositional mantle may -- on a geologically
short time scale (whose duration depends on both mantle thickness and
its thermophysical properties) -- become quickly charged with ice. The
results of this analysis bave important implications for understanding
the present distribution of ice within the near-surface crust and its
potential geomolphic expression.

14.03 Zent A. P.* Quinn R. C. Jakosky B. M.
Fractionation of Nitrogen Isotopes on Mars: The Role of the
Regolith as a Buffer

We have measured the adsorption of molecular nitrogen (N(sub)2) S on
palagonite, and modeled the adsorbed nitrogen inventory on the martian
regolith. We were motivated by the fact that models of isotopic
evolution predict stronger N2 fractionation than reported by Viking.
Possible scenarios for reconciling models with the observation include:
a heavy C0(sub)2 atmosphere early in the planets' history; continued
outgassing of N(sub)2 throughout the history of Mars; or a substantial
adsorbed inventory. We are investigating the plausibility of the latter
explanation. The adsorption measurements were made via a standard
volumetric adsorption apparatus. We modeled the adsorption on the
martian regolith by dividing the regolith into chunks, bounded every ten
degrees of latitude, and with 10 depth steps. The partial pressure of
N(sub)2- in the atmosphere E (16 Pa) is assumed to apply throughout the
regolith, and the adsorptive coverage of each chunk is calculated
separately and summed over the entire regolith. We parameterize the
regolith in terms of its total depth and specific surface area; we
consider specific surface areas between 10 and 100 m^2 g^-l, and
regolith depths from 50 to 500 m. If fractionation models are to be
reconciled with the observed fractionation solely through the presence
of 7 an adsorbed N2 component, we a calculate that approximately 59 kg
m^-2 of N(sub)2 must be adsorbed. We find that the regolilh reservoir of
adsorbed N(sub)2 is less than 50 kg m^-2, and so inadequate to buffer
the atmospheric isotopic composition. The most plausible explanations of
the observed isotopic fractionation require that N(sub)2 have outgassed
throughout martian history, or that early loss of N(sub)2 occurred at a
lower rate, the consequence of a more massive atmosphere. There is
supportive evidence for both explanations.

14.04 Lee S. W.* Clancy R. T. Martin T. Z.
Spatial and Temporal Mapping of Regional Surface Albedo and
Atmospheric Dust Opacity on Mars

Past studies have demonstrated that aeolian processes are active over
much of the surface of Mars; variations in regional albedo are
indicative of active sediment transport in a region, while thermal
inertia data are indicative of the degree of surface mantling by dust
deposits. The visual and thermal data are therefore diagnostic of
whether net erosion or deposition of dust-storm fallout is taking place
at present and whether such processes have been active in a region over
the long term. The earlier investigations have not corrected for the
effects of atmospheric dust loading on observations of the martian
surface, so quantitative studies of current sediment transport rates
have included large elTors due to uncertainty in the magnitude of this
"atmospheric component" of the observations.

A radiative transfer model has been developed which allows corrections
for the effects of atmospheric dust loading on observations of surface
albedo to be made. This approach to determining "dust-corrected surface
albedo" incorporates variable atmospheric dust opacity, the most recent
dust particle properties (Clancy et al, JGR 1994, in press), and
accounts for the lighting and viewing geometry of individual
observations. The spatial and temporal variability of atmospheric dust
opacity strongly influences the modeling results; we are making use of
the method developed by T. Z. Martin (JGR 84, 10941-10949,1993) to
determine dust opacity irom Viking IRTM thermal observations. This
modeling approach allows the atmospheric dust opacity to be determined
at the highest spatial and temporal resolution supported by the IRTM
mapping data; maps of "dust-corrected surface albedo" and atmospheric
opacity can be constructed at a variety of times for selected regions.
Information on the spatial and temporal variability of both surface
albedo and atmospheric opacity, and inferences of the amount of dust
deposition/erosion related to such variability, result. Analyses of
Viking IRTM mapping observations of the Syrtis Major region, spanning
more than one martian year, will be presented.

14.05 Sunshine J. M.* Mustard J. F.
Quantification of Variations in the Mafic Mineralogy of Mars
Through MGM Analysis of ISM Spectra

Absorptions near 1 and 2 micrometers in ISM imaging spectrometer data
(0.77 to 3.14 Sm) indicate that calcic-pyroxenes exist in weakly altered
volcanic terrains on the martian surface. Nili Patera (in Syrtis Major)
and Eos Chasma (in Valles Marineris) exhibit distinct differences in
absorption strength and shape, suggesting mineralogic differences. Here,
we report on new efforts to place limits on these possible compositional
variations by applying the Modified Gaussian Model (MGM) to ISM spectra.
The MGM is a method for deconvolving spectra into their constituent
absorption bands (Sunshine et al. JGR, 1990). However, because the
absolptions in ISM spectra are relatively weak (5-15%), modeling efforts
must be constrained through consistency with existing data bases and by
incorporating experience with MGM analysis of laboratory spectra of
pyroxene mixtures. Our initial goal was to test if the absorptions in
ISM spectra are due to a single pyroxene, or to two?? To explore this
hypothesis, we begin with simple models, making the fewest assumptions
about constituents, and then proceed, as the data warrant, to more
complex models. Examination of solutions in the context of laboratory
data, focusing on critical comparisons between the known coupling in the
1 and 2 micrometer regions, have led to the conclusion that both the
Nili and Eos regions contain two distinct pyroxene phases, a high- and a
low-calcium pyroxene, as well as a ferric component. No compelling
spectral evidence for an olivine was found. The differences in modal
mineralogy between Nili and Eos have been quantified as a 20% increase
in high-calcium pyroxene at Nili relative to Eos. MGM analysis of
spectra from more altered regions near Nili and Eos have revealed
quantifiable increases in the ferric component, while maintaining the
differences in the modal ferrous mineralogy. Thus, we have demonstrated
that the MGM can be used to examine crustal composition even through the
veil of dust on Mars.

14.06 Merenyi E.* Calvin W. M. Singer R. B.
Tracking Compositional and Physical Similarities from the
Martian Equator to the South Polar Region

In earlier works we characterized a previously unreported Martiar soil
type that covers Deucalionis Regio (centered at [245 degrees W,15
degrees S]) and also appears in part of Noachis, 20 degrees W - 340
degrees W, 30 degrees S - 50 degrres C (Merenyi et al. under review,
Icarus, Merenyi et al. PLPSC XXV). From analysis on a high spectral
resolution telescopic image in the 0.4 -1.02 micrometer range (Singer et
al. PLPSC , XXI), and on Viking IRTM data we concluded that this soil
was compositionally and physically significantly different from the
surrounding major surface units such as Arabia, Sinus Meridiani,
Acidalium, Chryse, and that it was likely to be immobile and possibly
cemented by salts. In a second telescopic image cube a south polar area
was found where the spectra were virtually identical to those of the
Deucalionis spectra longward of 0.7 micrometers. Based on works by
Kaufman (1989), Christensen and Zurel JGR, 1984), Iwasaki et al. (1990)
and others, the nature of the spectral deviation.under 0.7 micrometers
suggested the sout polar spectra migh be seen through a thin layer of
condensates. Preliminary analysis using the newly derived thermal
inertia map for the south polar region by Paige and Keagan ( JGR, 1994)
supported that Deucalioni and the S. Pole areas also show similar
physical properties. The compositional and physical likeness raised the
question of a possible genetic relationship between equatorial and south
polar regions spanning the southern hemisphere of Mars.

Here we present a more careful and detailed analysis that verifies the
above spectral and physical similarities. Through rigorous Mi modeling,
we show that the spectra from the south polar area described above can
indeed be Deucalionis type spectra modified by atmospheric scattering.
The possibility of an indurated surface is confirmed from the data of
Paige and Keagan with more refined discussion regarding the comparison
with mid-latitude thermal inertia maps and interpretation of the thermal
inertia values.

The existence of these similarities linking Deucalions, Noachis and some
areas on the south polar region would provide a piece of evidence in the
emerging picture of a global Martian compositional and morphological

14.07 Blaney D. L.* Crisp D. Martin T. Z.
Analysis of the Mariner 6/7 Infrared Spectra of Mars

The Mariner 6/7 Infrared Spectrometer experiment measured IR spectra of
Mars between 1.9 - 14.4 llm. These data were recently restored, with
improved radiometric and wavelength calibration and updated geometry
[1]. Although collected in 1969, they are still the only spacecraft data
available between 3.1 - 5 micrometers. The spacecraft6 tracks contain
data from latitudes ranging from roughly + 20 degrees N to -80 degrees
S, covering most of the geologic unit types on Mars. Spatial resolution
was about 130 km at closest approach and spectral resolution was roughly

We are reanalyzing this data set using modern atmospheric modeling
techniques to remove atmospheric gas and dust effects in order to
identify variations in surface mineralogy. A major concern in the
earlier analysis of the Mariner 6/7 data set by Pimentel et al. 1974 [2]
are differences in viewing geometry, dust opacity, and atmospheric
absorption features. We are currently focusing on the spatial
distribution of the 3 micrometers band caused by hydrated minerals on
Mars Our first step has been to try and recreate the results of Pimentel
et al. on the latitudinal dependance on the depth of the 3 ,micrometer
band. Although our calibration differs due to use of different solar
spectra, we still observe the same trends in the data the Pimentel et
al. see. The most hydrated regions occur at latitudes > 50 degrees S and
in the equatorial bright region while the least hydrated region was in
Meridiani Sinus (lat -4.2 degrees S, 359 degrees E). Detailed modeling
of the effects of the 2.75 micrometers CO2 absorption feature and of
atmospheric dust, which may also have a 3 micrometer absorption feature,
will be done to fully characterize the distribution of hydrates on Mars.

1. T.Z. Martin, Restoration ofthe Mariner 6/7 IR Spectrometer Data Set.
Icarus,in press, 1994; 2. G.C. Pimentel et al. J. Geophys. Res. 11, 1
623,1 974. This work was done at JPL/ CALTECH, under contract to NASA.

14.08 Hansen G. B.*
Low Brightness Temperatures of Martian Polar Deposits: New
Insights from Recent Measurements of CO2 Ice

The infrared thermal mapper (IRTM) instruments aboard both Viking
orbiters made the first extensive measurements of infrared radiance
during the Martian polar night. Early measurements of the south polar
region in polar night showed brightness temperatures in the 20
micrometers band well below tne expected temperature of CO(sub)2 frost
deposits (Kieffer, et al. (1976), Science, 193, 780; (1977), J.
Geo-phys. Res., 82, 4249) Many explanations have been advanced for this
behavior, including the low emissivity of the frost deposits, a position
supported by the contemporary measurements of Ditteon and Kieffer
((1979), J. Geophys. Res., 84, 8289). Paige ((1985), PhD Thesis, Calif.
Inst. of Tech., Pasadena, CA) used information from the 11 micrometers
channel as well as the 20 micrometers channel in a detailed study of the
polar night measurements in both the north and south, and called into
question the emissivity argument due to the spatial and temporal
variation of the two brightness temperatures.

New measurements of the spectral transmission of CO(sub)2 ice indicate
that in the spectral regions on either side of the strong l5 micrometers
absorption, it is at least a factor of 10 times more transparent than
reported by Ditteon and Kieffer. Furthermore, the absorption is higher
across the 11 micrometers band than across the 20 micrometers band,
indicating that the emissivity (and brightness temperature) will be
higher in the 11 micrometers band, and moreover, that the relative
emissivities in the two bands will vary with effective particle size. A
radiative transfer model of pure CO(sub)2 frost with kinetic temperature
146 K gives the following nadir brightness temperatures (T(sub)B).

particle dia. T(sub)B ( 1 1 llm ) T(sub)B ( 20 micrometers )
10 micrometers 119 . 2 K 97. 4 K
100 micrometers 130.8 K 111.3 K
1 mm 140.0 K 126.6 K
10 mm 144.3 K 139.0 K

The addition of small amounts of water ice or dust raises the brightness
temperatures considerably and may change the relative temperatures
between the two bands. By varying the water ice and dust content and
particle size, the observed spatial and temporal variations of brightess
temperatures may be simulated.

This work was supported JPL, Caltech, and NASA.

14.09 Slade M. A.* Jurgens R. F.
Goldstone Radar Observations of the 1995 Mars Opposition

As part of the International Mars Watch, Goldstone radar observations of
Mars are planned during the 1995 Opposition (~Dec.'94-March'95). While
much of the allocated time is already devoted to observations for Mars
Pathfinder landing site validation, a small number of tracks could be
made available for well-focused scientific objectives. During this
opposition, the sub-Earth latitudes are in Mars' Northern hemisphere
(Figure 1,) over terrain not previously examined with current
sensitivity. The radar data types possible in 1995 will be described at
he meeting. A good overview can be found in Chapter 20 of the book Mars
(Keiffer et al., ed.; 1992).

Informal observing proposals from any scientist with peer-reviewed Mars
funding should be forwarded to Martin.A.Slade@jpl.nasa.gov by email.
Data reduction can, in principle, be carried out over the Internet. A
graduate student or postdoctoral fellow resident at JPL for short period
is recommended, however, to become familiar with suite of software for
data analysis. (Unfortunately, JPL cannot supply travel reimbursement
due to the current NASA funding situation.) The Mars 1995 dataset should
provide a significant increase in the knowledge about the surface and
subsurface of Mars. We urge that you consider becoming involved with the
acquisition and analysis of Mars radar data.

Figure 1 appears here in the hard copy.

14.10 Pischel R.* de Niem D.
Classification of Surface Features from Mars '96 HRSC and WAOSS
Imagery Using Statistical Analysis and Image Processing Methods

The analysis of surface features plays a key role for the understanding
of the geology and the evolution of planets and small bodies.
Statistical analysis methods are most important for the interpretation
of geological structures. Digital image analysis approaches are used for
systematic processing steps and interpretation suppport in order to
handle huge data volumes as they are expected for the High Resolution
Stereo Camera (HRSC) and Wide Angle Optoelectronic Stereo Scanner
(WAOSS, imagery on the Mars '96 mission (approximately 200 GByte
decom-pressed data).

One of the main features of the HRSC and WAOSS cameras is there stereo
capability. Images with a planimetric resolution of 10-20 m and an
absolute height accuracy of 70 m are expected from the high resolution
(HRSC) data. The goal of this study is to investigate how this HRSC high
resolution image data and HRSC and WAOSS stereo image data can be used
for the analysis of selected Martian surface features based on a
combination of statistical methods and a pattern recognition approach.
The use of this technique is shown for the case of crater detection and

An approximation approach is proposed which is applied to a combined
digital image and digital elevation data set. A number of molphological
features (crater form, peak height etc.) and crater parameters can be
easily derived using the "approximated crater". One example for
statistical analysis methods is the succeeding classification step which
is based on feature information (e.g. existence of inner peaks) as well
as on numerical data (e.g. curvature of the crater rim).

Results are presented using Viking images and simulated data.

14.11 Stoker C.*
Field Experiments Using Telepresence and Virtual Reality to
Control Remote Vehicles: Application to Mars Rover Missions

This paper will describe a series of field experiments to develop and
demonstrate the use of Telepresence and Virtual Reality systems for
controlling rover vehicles on planetary surfaces. In 1993, NASA Ames
deployed a Telepresence-Controlled Remotely Operated underwater Vehicle
(TROV) into an ice-covered sea environment in Antarctica. The goal of
the mission was to perform scientific exploration of an unknown
environment using a remote vehicle with telepresence and virtual reality
as a user interface. The vehicle was operated both locally, from above a
dive hole in the ice through which it was launched, and remotely over a
satellite communications link from a control room at NASA's Ames
Research center, for over two months. Remote control used a
bi-directional Internet link to the vehicle control computer. The
operator viewed live stereo video from the TROV along with a
computer-generated graphic representation of the underwater terrain
showing the vehicle state and other related information. The actual
vehicle could be driven either from within the virtual environment or
through a telepresence interface.

In March 1994, a second field experiment was performed in which the
remote control system developed for the Antarctic TROV mission was used
to control the Russian Marsokhod Rover, an advanced planetary surface
rover intended for launch in 1998. Marsokhod consists of a 6-wheel
chassis and is capable of traversing several kilometers of terrain each
day. The rover can be controlled remotely, but is also capable of
performing autonomous traverses. The rover was outfitted with a
manipulator arm capable of deploying a small instrument, collecting soil
samples, etc. The Marsokhod rover was deployed at Amboy Crater in the
Mojave desert, a Mars analog site, and controlled remotely from Los
Angeles in two operating modes: (1) a Mars rover mission simulation with
long time delay and (2) a Lunar rover mission simulation with live
action video. A team of planetary geologists participated in the mission
simulation. The scientific goal of the science mission was to determine
what could be learned about the geologic context of the site using the
capabilities of imaging and mobility provided by the Marsokhod system in
these two modes of operation.

I will discuss the lessons learned from these experiments in terms of
the strategy for performing Mars surface exploration using rovers. This
research is supported by the Solar System Exploration Exobiology,
Geology, and Advanced Technology programs.

14.12 Paige D. A.*
The Mars Polar Pathfinder Discovery Mission

The Mars Polar Pathfinder is a Discovery-class mission whose primary
scientific objective is to initiate the exploration of the Martian north
polar ice deposits. On Earth, polar ice deposits have yielded detailed,
continuous records of past climatic variability extending from the
present to hundreds of thousands of years into the past. On Mars there
is a strong theoretical basis for expecting significant
astronomicaily-driven climate variations, and there is strong
observational evidence that Mars experiences significant interannual
climate variability as well. If the Martian polar ice caps and
associated layered deposits are also found to contain records of past
climatic variability, they will surely become major focal points for the
scientific exploration of Mars in future decades. The Mars Polar
Pathfinder is intended to take the first steps towards this ultimate
goal by characterizing the basic composition and stratigraphy of the
Martian north polar ice cap at a single landing site.

As the name suggests, the Mars Polar Pathfinder Project derives
considerable heritage from Mars Pathfinder. It employs the same Delta
launch vehicle, and uses a close copy of the JPL-built Mars Pathfinder
lander. The mission plan calls for launch in August 1998, and landing
near the Martian north pole in November, 2000, which is just after the
start of the Martian northern summer season. Once landed, the spacecraft
will deploy its instruments and operate on the surface of the north
polar water ice cap using solar power and direct to-earth communications
for approximately 100 days.

The strawman scientific payload for the Mars Polar Pathfinder includes:
a descent imager, a meteorology package and a surface imager which
together will survey the local environment at the landing site; an
ice-sounding radar which will determine the internal structure and total
depths of the polar ice deposits; a drill and borescope system, a
robotic arm and ice/dust thermal analyzer which together will determine
the basic composition and stratigraphy of near-surface layers. Other
instruments being considered for inclusion on the payload include a
thermal probe, a micro mass spectrometer, and an extraterrestrial
resource utilization demonstration experiment.

The Mars Polar Pathfinder can achieve its exciting scientific and
mission objectives within the cost guidelines of the Discovery Program
primarily because of its extensive heritage from the Mars Pathfinder

14.13-T Hartmann W. K. Engel S.
Studies of Ancient Mars and Titan Atmospheres Using the
Cratering Record

We find that there may be significant opportunities to study the past
history of the atmospheres of Mars and Titan by using the cratering
record on surfaces of different ages.

Using a program originally developed by Chyba, Thomas, and Zahnle
(1993), we have calculated minimum sizes of bolides that can make it
through present atmospheres of Mars and Titan without exploding. This
filtering produces an observable turndown in the crater diameter
distribution on these bodies. Similar calculations have been made by
Melosh and others for the present-day atmosphere, but our work differs
by assuming higher past atmospheric densities to examine at what sizes
the turndown occur for hypothetical ancient atmospheres of different

For Mars, we conclude that if the atmosphere was ever denser than about
200-300 mb, cometary bolides and chondritic stony asteroids would be
filtered by the atmosphere, producing a strong downturn in the cratering
record near crater diameter 0.5-4 km (resembling that observed on Venus,
but at smaller sizes). Only irons would make smaller craters.

Titan has a massive atmosphere with a high scale height, and we find
that it is comparable to the atmosphere of Venus in filtering effects.
The smallest cometary bodies that would get through Titan's present
atmosphere would produce a minimum crater size of tens of kilometers.
This effect should be detected by Cassini radar. Competing theories
picture either a denser or less dense ancient Titan atmosphere; tests of
these theories may be possible, depending on the erosion state of Titan.

Chyba, C., Thomas, P., and Zahnle, K. (1993) Nature 361, 40.

14.14-T Bell J. F. III Bornhoeft K. Lucey P. G. Pollack J. B.
High Resolution 0.50 to 0.95 Micrometer Spectroscopy of Mars
During 1990

We conducted visible to near-IR imaging spectroscopic observations of
Mars from Mauna Kea Observatory during the 1990 opposition. Our goal was
to complete imaging spectroscopic survey begun in 1988 by imaging
regions of the planet centered the Amazonis-Elysium hemisphere not
studied previously. The data were obtained on the University of Hawaii
2.2-m telescope using the Wide Field Grism Spectrogra instrument. We
obtained approximately 10,000 spectra from 0.50 to 0.95 micrometers of
regions of the surface visible on 9 and 10 November 1990 UT. The spatial
scale of the data was 0.7 arcsec/pixel, corresponding to a resolution on
Mars of 540 km/pixel at the sub-Earth point. Seeing conditions were not
optimal, and thus the imaging aspect of this experiment was not very
successful. We have concentrated on the interpretations of the spectra
instead because they exhibit high S/N and reveal diagnostic absorption

Telluric atmospheric correction and solar spectral removal were achieved
using the Hardorp solar analog star Kappa Cet (G5V). The Mars / Kappa
Cet spectra (~ Mars / Sun) show interesting features that support and
enhance interpretations gleaned from previous observations. These
features include: (1) a steep red 0.50 to 0.75 micrometers spectral
slope both bright and dark regions, with a bright/dark ratio near unity
at 0.50 micrometers and increasing to 1.5 to 2.0 near 0.75 micrometers;
(2) a strong (for Mars) absorption feature in the 0.8 to 0.9 micrometers
region, center near 0.86 micrometers in bright regions and near 0.89 in
darker regions; (3) a weaker, partial absorption band or "cusp" in the
spectra of both bright and dark regions, centered near 0.68 micrometers;
and (4) a distinct difference between the near-lR spectral slope of dark
regions (negative) and bright regions (flat).

Our interpretation of these spectra is that the absorption band
positions in the bright regions indicate the presence of a small arnount
(less than 5%) of well-crystalline hematite that must occur in a matrix
of much more poorly crystalline ferric-bearing materials which account
for the overall color of Mars. The shift in band position observed
between bright and dark regions and the negative 0.80 to 0.95
micrometers spectral slope are most likely caused by the presence of
pyroxene-bearing materials in the dark regions that have ferrous
absorption band positions longward of 0.9 micrometers. An alternate
hypothesis, that the band shift is due to the presence of hydrous ferric
oxides, is possible but less likely given the position of the near-UV
absorption edge and the decreasing near-lR spectral slope.

Because we observed a hemisphere dominated by bright, windblown
materials and our spatial resolution was poor, there is certain to be a
high amount of actual and effective spatial mixing occurring in the
data, such that the spectral differences between regions are muted and
even the darkest spectra measured have a substantial "bright" component
to them. The only solution to this problem is increased spatial
resolution, either utilizing improvements in groundbased techniques,
spaceborne observatories, or Mars-orbiting spacecraft. All of these
approaches will be attempted by many different groups over the next
decade, which will greatly improve our understanding of Mars spectral

14.15-P Erard S.
Variability of the Surface Materials in the Equatorial Regions
of Mars

A systematic search for spectral variations was performed on imaging
spectroscopic data of Mars acquired by ISM onboard Phobos-2. The dataset
consists in nine observation sessions, each comprised of about 3,000
spectra of pixels 25 x 25 km^2 in size, from 0.77 to 3.15 micrometers.
Following preliminary analysis of th the spectra were described by 5 to
7 spectral parameters as less sensitive as to photometric effects and
atmospheric scattering. Those parameters are mostly related to iron
mineralogy and hydration. They were checked on possible martian analogs
to insure they allow characterization of mineralogical composition.

The nine obselvation sessions were studied independently by means of
analysis, with relatively high security levels. The analysis yielded 4-7
main units per session, some of them being macroscopic mixtures of pure
end members. The corresponding spectral types were then used as input to
a new analysis so as to get a global image of spectral variability in
the observed regions. This study confirms the higher variability of dark
regions and the diversity of bright materials, mostly related to the
presence of different iron oxides. Current investigations focused on the
interpretation of the materials in terms of mineralogical/chemical
composition. This work was partly supported by an ESA fellowship.

14.16-P Cohen B. A. Singer R. B.
Detectability of Siderite (FeCO3) in Martian Soil: An
Experimental Investigation

It has been postulated that the Martian surface could contain abundant
carbonates [1,2], based on evidence that Mars once had both a
substantial CO2 atmosphere and liquid water reservoirs. While only minor
carbonate has been detected on the whole surface [3], buried carbonate
sediments may be only locally exposed and observable with high spatial
resolution sensors. If Mars did once support liquid water under a
reducing CO2 atmosphere, metal ions would have been weathered from young
volcanic rocks but would not have been rapidly oxidized as on Earth.
This might have led to a saline, acidic ocean, where calcite (CaC03) and
siderite (FeCO3) would form. Recently, the abundance of iron oxides on
the surface has led to speculation that siderite could be at least as
important as calcite [4]. This study examines the VIS-NIR reflectance
spectra of siderite mixed with Martian surface analogs. The reflectance
spectroscopic properties of calcite both by itself and in mixtures with
palagonite have been investigated [5,6]. In addition to IR vibrational
features, siderite has a very prominent, broad double band at about 1
micrometer [7]. Calcite and palagonite are relatively unabsorbing here.
Samples used were natural siderite, palagonite, and calcite,
mechanically mixed together. Analysis was performed from ,0.33 to 1.2
micrometers with the PIRL laboratory spectrophotometer (modified from a
CARY-14). Both the depth and position of the l-micrometer band were
found to shift regularly with increasing amounts of siderite. Mixtures
of 20 wt% siderite with varying amounts of calcite and palagonite also
exhibit good correlations between band depth/center and the percent
CaC0(sub)3 in the mix. Experimentally, it is shown that siderite could
be detected in a palagonite-like Martian soil in quantities as low as 6
wt% if there is no calcite present. A more likely scenario would be a
carbonate deposit containing both iron and calcium carbonates; in this
situation, even smaller amounts of siderite would be detectable, as
siderite's l-micrometer band remains visible at much lower abundances
when calcite is present. Vibrational bands are diagnostic of carbonate
minerals, but VIS-NIR spectra are useful in making cation distinctions,
and have potential for estimating the proportions of each phase. The IMP
camera on Mars Pathfinder is sensitive to these features.

REFERENCES: [1]Gooding (1978) Icarus 33, p.483. [2]Clark and VanHart
(1981) Icarus 45, p.370. [3]Roush, et al. (1989) LPSCXX, p.928 [4]Banin
et al. in Mars, ed. Kieffer et al. ( 1992: U of A Press) p. 594
[5]Blaney et al. ( 1987) LPSC XVIII, p. 87. [6]Wenrich and
Christensen(1994)LPSC XXV,.p.1485 [7] Gaffey(1987)JGR 92 B2,p.l429.

14.17-P Barlow N. G.
Impact Crater Depth-Diameter Ratios in Mangala Valles, Mars

The author is continuing studies of the variations in impact crater
depth- diameter ratios in selected regions of Mars. Results from the
photoclinometric studies have previously been reported for the Arabia,
Maja Valles, and Memnonia regions of Mars. The study has now been
expanded to the Central Mangala Valles region, extending from 7S to
14.5S and 148W to 160W. Forty-five Viking Orbiter frames with
resolutions between 45 and 50 m/px have been utilized in this analysis.
The Davis and Soderblom photoclinometry technique, specifically designed
for use with Viking Orbiter images of Mars, was used to determine crater
depths and diameters for all craters >k km in diameter on the images.
The depth-diameter ratios of fresh impact craters (identified by their
morphologically fresh appearance, such as pristine ejecta blankets,
sharp rim crests, and lack of obvious floor deposits) were determined
based on the results of the photoclinometric analysis. The
photoclinometric results were compared with results from shadow
measurements--the two techniques provided results within 2% of each

Simple craters (crater with diameters generally <6 km) within the
Mangala Valles region of Mars have an average depth-diameter ratio of
0.20 (+/- 0.05). Complex craters display an average depth-diameter ratio
of 0.08 (+/- 0.05). These ratios are similar to those obtained by this
method for fresh impact craters in the Arabia and Maja Valles regions.
Thus, the Medusae Fossae deposits in the Memnonia region are currently
the only area where anomalous depth-diameter ratios have been

14.18-P Hartmann W. K. Engel S. Chyba C. Sagan C.
Mars Cratering Record as a Probe of Ancient Pressure Variations

There are significant opportunities to study the past history of the
atmosphere of Mars by analyzing the cratering record on surfaces of
different ages [1]. As shown by the case of Venus, a dense enough
atmosphere leaves a dramatic cutoff in the cratering record below a
critical size. Dense as well as episodic ancient martian atmospheres
(>>10 mb), proposed by many authors, could cause similar effects.

Using a program originally developed by [2] and modified for Mars, we
have calculated minimum sizes of bolides that can make it through
present and hypothetical denser past atmospheres of Mars without
exploding. This filtering produces observable downturn in the crater
diameter distribution. Although calculations have been made by [3] for
the present atmosphere, our work calculates the effect for various
possible atmospheric densities that may have existed in the past.

We conclude that if the atmosphere ever reached more than about 100-300
mb surface pressure, cometary bolides and chondritic stony asteroids
would be filtered out by the atmosphere, generating a strong downturn in
the cratering record near crater diameter 0.5-4 km (resembling that
observed on Venus, but at smaller sizes). Only irons would get through
to make smaller craters (representing only about 1% to 4% of the craters
that would exist otherwise at small diameters).

In the course of our work, we have studied examples of isolated crater
clusters on Mars (typically tens of craters, D~700 to 1000 m). These are
important because they may identify the threshold of breakup for weak
cometary bodies--consistent with models of comet P/SL-9.

References: [1] Chyba C. et al. (1992) BAAS, 24, 975. [2] Chyba C. et
al. (1993) Nature, 361, 40. [3] Melosh H. J. (1989) Impact Cratering: A
Geological Process, Oxford University.

14.19-P Savijarvi H.
Thermal Inertia on Viking Sites in Mars

The thermal inertia of the uppermost martian regolith has traditionally
been determined indirectly by fitting surface temperatures predicted by
thermal models to remotely sensed surface brightness temperatures. The
soil models and their atmospheric forcing have been simplified and the
surface temperatures derived from the brightness temperatures contain
errors from various sources.

We have recently coupled our dusty and moist Helsinki Mars Mesoscale
Model (providing reasonable atmospheric forcing to the surface, [1,2])
with a numerically accurate regolith heat diffusion scheme [3]. By
adjusting the model-produced temperatures to the direct Viking 1.6 m
observations, an estimate of the ground thermal inertia is obtained as
the function of reqolith depth.

References: [1] Savijarvi H., Contr. Atmos. Phys., 64, 219-229. [2]
Savijarvi H. and Siili T., J. Atmos. Sci., 50, 77-88. [3] Savijarvi H.,
Contr. Atmos. Phys., 65, 281-292.

14.20-P Neugebauer G. Machling M. Dorrer E. Zhou X. Hauber E.
Hoffmann H. Jaumann R. Neukum G.
Cartographic Representation of the Martian Surface: New Results
for Digital Deshading and Reshading

In preparation for map production using data from the forthcoming camera
experiments onboard the Russian Mars'94/'96 missions (High Resolution
Stereo Camera and Wide Angle Optoelectronic Stereo Scanner), a
preliminary topographic image map of Tithonium Chasma, in the Valles
Marineris, has been produced. Such maps are supposed to show all
characteristics of a full topographic map, such as optimal true color
representation, relief visualization by prevailing NW-illumination,
appropriate topographic information (contour lines), map annotation, and

The input used to prepare the first experimental output in A4 format
consisted of two sets of information: (1) a 1/500,000 topographic
orthophoto mosaic (USGS 1980, Miscellaneous Investigations Series, Map
I-1294); (2) monochromatic image data from the Viking Orbiter 1 mission
with a pixel size of 60 m. The 200m contour lines of the topographic
orthophoto mosaic were digitized and generalized with an equidistance of
500 m. A digital elevation model (DEM) was derived from the 200m contour
lines. The image data were then mosaicked and co-registered to the DEM
to match the map geometry. Since the azimuth and elevation of the Sun
are known for the images, shadows could be simulated on the DEM and then
removed, as far as possible, from the image mosaic. According to
cartographic standards, the DEM was illuminated with a prevailing
NW-illumination. In order to improve the impression of the surface
relief, secondary illumination from different directions was added for
NW-SE striking morphological features.

In this preliminary work, color has been visually introduced to the DEM
using Viking Lander data in order to produce the Cyan-Magenta-Yellow
color separates for the printing process. However, it is planned to use
color data from Viking Orbiter to generate near true color
representation of the martian surface. During the map production, the
image information has been printed as the black channel (K) in the CMYK
model. In different experiments, the percentages of CMYK values have
been varied to optimize the visibility of the superimposed
contour lines as well as the underlying image mosaic.__

14.21-P Wells W. K. Sprague A. L. Hunten D. M. Rizk B. Hill R. E.
Mars: Observations of Seasonal Atmospheric Water Vapor

Ground-based high-resolution spectroscopic observations of water vapor
in Mars atmosphere have been made using the 8189 Angstrom line. Strong
seasonal variations are seen in the frames reduced to date Abundances at
all latitudes are similar to those seen by the Viking MAWD experiment
[1] and by Rizk et al. [2]. A simple latitudinal binning into northern,
equatorial and southern regions shows strong seasonal variations. Summer
values are a factor of two higher in the northern hemisphere than in the
southern hemisphere. Nineteen nights of good data were obtained. Below
are plotted the results of the eight reduced so far.

References: [1]Jakosky and Farmer, JGR, 87, 2099, 1982. [2] Rizk et al.
Icarus 9O 205, 1991

14.22-P Mumma M. J. DiSanti M. A.
Atmospheric Water Vapor on Mars: A Search for Enrichment of
HDO/H20 in the North Polar Cap

We present latitudinally resolved spectra of Martian H20 and HD0, which
permit us to address the question of whether HD0 is enriched in the
polar ice caps.

Existing models developed for the odd-hydrogen chemical cycle on Mars
have neglected the role of transport within the atmosphere. We have been
investigating this problem with the use of spatially-resolved global
maps of C0, H20, HD0, and C degrees 2 absorption lines, to assess
transport and photochemistry in the Martian atmosphere. We observed Mars
UT 9 - 12 March 1993, using CSHELL (resolving power lambda \delta lambda
~2 x 10^4) at the IRTF. At this time L(sub)s 50, degrees or northern
spring; thus the north polar cap would have just begun to sublimate.
Based on Viking Orbiter MAWD measurements, virtually no water vapor is
expected for southern latitudes, for this L(sub)s-value (Jakosky and
Farmer 1982 JGR 87, 2999).

A principal advantage in using modern array detectors such as CSHELL is
that they provide simultaneous spatial coverage for extended objects. We
onented the slit north-south, allowing full latitudinal coverage over
the disk of Mars, and targeted the H20 111-000 band at 8820 cm^-l, and
the HDO Vl band at 2720 cm^-l. Using spectra of the southernmost
latitudes as calibrators, we present latitudinal maps of H20, and of
HDO, which allow us to determine whether HDO is preferentially
sequestered in the northern polar ice cap reservoir.

14.23-P West F. R.
An Investigation of Advantages of Observing Mars from
Antarctica in 1997, 1999, 2001, 2003, and 2005

Advantages of observing Mars from Antarctica near and during the
upcoming opposition of 1997, 1999, 2001, 2003, and 2005 are
investigated. Several Antarctic sites, including 90 degrees S (South
Pole) and 82 degrees S (hig plateau) are investigated for the visibility
of Mars above the horizon during these years. An important
characteristic investigated is the possibility for keeping Mars under
continuous observation for at least several weeks during Antarctic
autumn and winter (March 21 through Sept. 23). Detailed calculations of
periods when Mars is continuously above the horizon for more than a day
and for other possibly favorable observing periods are summarized for
the chosen observation sites.

Projects that could benefit from prolonged continuous monitoring of Mars
are studies of dust storm development and changes in its polar regions.

Preliminary studies indicate that the 1997 and 2005 oppositions will not
be favorable for such an observational project from Antarctica, but that
the oppositions of 1999 and 2001 will be quite favorable. The opposition
of 2003 promises to be excellent.

SESSION 15 ....... Comets I
Wednesday, 8:30 - 10:00 Crystal Ballroom A
D. C. Boice and R. Meier, Moderators

15.01 Rodin A. V.*
Vibration and Rotational Distribution of C2 in the Inner Coma
of P/Halley: Update Analysis of the Results of VEGA Mission

Emission in the bands Delta v = -1,0,1 of C(sub)2 Swan system measured
by means of TKS spectrometer with scanning field of view during VEGA
flyby has been analysed using line-by-line model involving 10 vibration
levels and J up to 60. Inversion of low-resolution data on the entire
band sequence permits to retrieve smoothed rotational structure of v' =
0 state as well as vibration population with an accuracy sufficient to
reveal features specific for newborn molecules in the innermost fields
of coma not available for ground-based spectroscopy. Spectra observed
with line of sight impact parameter less than 300 km yield bimodal
Bolzman distribution with inflection point around J'=30 and
characteristic temperatures T(sub)l = 2730K and T(sub)2 = 5700 K. Since
distribution variations for J<10 are left out of analysis and cold
component cannot be model-independently retrieved, comparison with
laboratory data combined with spatial fitting map by gas outflow
modelling confirms an assumption that main source of C near the nucleus
is two-step photodissociation of acetylene. In the framework of
available accuracy, observed distribution relaxes to single Bolzman mode
with T = 3300K at nucleocentric distance about 1500 km, according to a
variety of models of C(sub)2 fluorescent equilibrium. However, quasi-3D
section of the coma shows that an extended source associated with jet
containing CHON particles contributes to radical production with thermal
rotational population and lower vibration excitation initial states of
remnants are probably controlled by evaporating phase.

15.02 Luu J.* Jewitt D.
Colors of Centaurs and Trans-Neptunian Objects

The planet-crossing orbits of the Centaurs are short-lived, implying
that they must have originated elsewhere in the solar system. The newly
discovered trans-Neptunian Kuiper Belt provides a possible source of the
Centaurs, and of short-period comets. In this paper we present
broad-band optical colors of the known Centaurs and of the brighter
trans-Neptunian objects. The colors represent our first attempt to
compare trans-Neptunians with each other, and with the three known
Centaurs. The observations are complicated by the faintness of the
trans-Neptunians (red magnitudes are in the range 22 to 24.5). Our main
result is that the trans-Neptunians possess diverse colors, and that the
striking red mantles of 5145 Pholus and 1993 HA2 are not representative
of all Kuiper belt objects. Possible implications for the processing of
surface materials on distant objects will be discussed.

15.03 Kolokolova L. O.*
Studying Cometary and Asteroid Composition Using Stokes
Parameters: Computations

A new successive reflections method for computing the Stokes parameters
light multiply scattered by rough surfaces and large fluffy particles
suggested. The method is based on the order-of-scattering approach
applied earlier to planetary atmospheres by Uesugi and Irvine (1970). We
have obtained a good quantitative agreement between theoretical
computations and laboratory measurements of the Stokes vector for powder
samples (Degtyare and Kolokolova, 1992). In comparing theoretical
computations and laboratory measurements, we analyzed data for all the
four Stokes parameter as functions of the phase angle. We have shown
that both linear and circular polarizations are sensitive to the
chemical composition and texture of scattering particles, and that the
third and fourth Stokes parameters are dependent on the particle shape
and presence of anisotropic features on the surface. The dependence of
the Stokes parameters on the refractive and absorption indices and
structural properties have also been analyzed. We make some suggestions
regarding the use of Stokes parameters in remote sensing of asteroid
surfaces and cometary particles.

References: Degtyarev V. and Koldlcolova L. (1992). Earth Moon Planets
34, 45. Uesugi A. and Irvine W. (1970). Astrophys. J. 159, 127.

15.04 Fomenkova M.* Larson S. Jones B. Pina R.
Albedo Map of Comet Swift-Tuttle

Comet Swift-Tuttle is an active short period comet. Quasisimultaneous
images of this comet were obtained on Nov 09, 1991 and on Nov 17,1991 in
the R-band (with the plate scale of 0.7"/pixel) and at 11.6 microns
(with the plate scale of 0.8"/pixel). These data were used to produce
albedo maps in accordance with [1]: gamma = 1/(1 + IR/VIS), where gamma
is albedo, IR is the thermal emission image, and VIS is the reflected
sunlight image.

On Nov 09, the comet was in a quiescent state. The albedo map covers the
10^4 by 10^4 km region of the coma. The value of the albedo is 0.03 at
the center of the coma and gradually increases with the distance from
the nucleus up to 0.2. On Nov 17, the map covers the 3 x 10^4 by 3 x
10^4 km region of the coma. The comet had two pronounced jets. The
albedo of the dust ranges from 0.05 up to 0.08 in the jets and from 0.12
up to 0.3 in the rest of the coma. The lowest albedo is in the downward
jet which is the closest to the sunward direction. The albedo of
particles in the upward jet is slightly higher.

This behavior is consistent with the hypothesis that fresh dust,
recently emitted from the nucleus, has on average lower albedo - due to
the presence of either larger particles or darker (presumably organic)
material, or both. After leaving the nucleus, the particles fragment or
the volatile material decomposes due to heating and solar radiation,
thus leading to an increase of the albedo.

References: [1] O'Dell C. R. (1971) Ap J 166, 675-681; Ney E. P. (1982)
In Comets (ed. L. Wilkening), pp. 323-340, U. of Arizona Press.

15.05 Jewitt D.* Luu J.
The Solar System Beyond Neptune

We are studying the trans-Neptunian solar system using deep imaging
observations at optical wavelengths. Distant objects are identified by
their slow, retrograde opposition motions. Thirteen trans-Neptunians
have been identified at the time of writing (June 1994), mostly as part
of a survey to magnitude 25 conducted on Mauna Kea. These are the first
detected members of a trans-Neptunian swarm that comprises about 35,000
objects larger than 100 km in the 30 - 50 AU heliocentlic distance
range. Vast numbers of smaller objects presumably exist below the limits
of detection.

The dynamical natures of the new objects are unclear. The best
characterized (1992 QB1, 1993 FW) have perihelia near 40 AU, and so are
relatively immune from strong Neptune perturbations (Neptune is at 30
AU). These objects, as well as the distant 1994 ES2, 1994 EV3 and 1994
JQl (heliocentric distances 46, 45 and 43 AU, respectively) are likely
members of the predicted Kuiper Belt. Other objects at smaller distances
(e.g. 1993 RO, RP, SB and SC, all near 32 - 35 AU) might be part of the
Kuiper Belt, or could also be stabilized by gravitational resonance with
Neptune either as trojans (1:1 resonance) or"Plutinos" (3:2 resonance).

We will discuss the observational status of the solar system beyond

15.06 Sykes M.* Cutri R. Moynihan P. Plath J.
A Parallactic Mini-Survey of the Infrared Sky

We are probing for distant members of the solar system in the Kuiper
Belt and inner Oort Cloud using IRAS images constructed from scans of
the sky at 60 and 100 micrometers. These would include hypothesized
cometary collision debris and "local" cirrus. Cirrus is seen at all
galactic latitudes and evidences structure at all spatial scales. It is
likely that the solar system passes through or by these clouds as it
orbits the galaxy. Parallactic distances probed by pairs of scans in the
case where a source is seen to shift 3' correspond to a depth of
1500-3000 AU when using the larger baseline observations. To search for
sources near the ecliptic we have developed a process by which survey
images are "cleaned" of structures such as the zodiacal dust bands.

15.07 Haken M.* A'Hearn M. F. Feldman P. D.
Secular Variation in Comet P/Encke

Comet P/Encke, with the shortest period (~3.3 yr.) of any known comet,
affords the best opportunity to study secular variation of brightness in
comets. We have observed P/Encke with IUE over five consecutive
apparitions, from 1980 to 1994, three times pre-perihelion and twice
post-perihelion. Comparison of pre-perihelion data from the first (1980)
and the last (1994) of the observed apparitions revealed no evidence of
secular change in the outgassing rate of water, based on the OH emission
at 3086 Angstroms. In particular, the secular fading, which some authors
have reported was not confirmed. The maximum such variation consistent
with our data is unclear, due to the fact that different cameras aboard
IUE were used for the two sets of observations, and there exists no
definitive assessment of the relative error in the absolute calibration
of the two cameras. However, we can certainly set an upper limit on
fading of 2.5% per revolution, and may be able to further constrain the
possible size of this effect. To do so, we will reduce and compare
spectra of standard stars taken with the two cameras to assess
systematic differences.

This work was supported by NASA's IUE program.

15.08 Eberhardt P.* Krankowsky D.
The Electron Temperature in the Inner Coma of Comet P/Halley

From the CH(sub)30H(sub)2^+/H(sub)30^+ and H(sub)30^+/H(sub)20^+ ratios
measured by the Giotto Neutral Mass Spectrometer in the inner coma
(distance < 20'000 km) of comet P/Halley we derive the electron
temperature profile. The electron temperature T(sub)e increases
gradually from ~ 100 K at 2000 km to approximately 10^3 K at 8500 km
followed by a sharp increase to well above 10^4 K at 11'500 km. Then
T(sub)e drops gradually to values below 10^4 K at 20'000 km. This
profile is used for modeling the absolute H(sub)20^+, H(sub)30^+, and
CH(sub)30H(sub)2^+ densities. We obtain excellent agreement between
modeled and measured ion densities over the whole distance range from
inside the contact surface to 20'000 km. Electron impact ionization
plays no important role in this region. The up to fourfold ion density
enhancement observed for distances between 10'000 and 20'000 km (pile-up
region) is the result of the sharp Te increase coupled with very low
radial ion velocities in this distance range.

15.09 Esterle P. A.* A'Hearn M. F.
Comet Halley's Rotation from Observations of Dust Jets from
Giotto and the Ground

There is much debate among cometary astronomers as to the 'correct'
rotational state of Comet P/Halley. Recently, two completely different
models--a Fast Long Axis Mode with periods of 2.2 and 3.6 days, and a
Slow Long Axis Mode with periods of 3.6 and 7.3 days--have been invoked
to describe many observed phenomena. The most probable of each of these
models have been tested against a series of ground-based and spacecraft
observations of dust jets in Comet P/Halley. While the Fast Long Axis
Mode can be used to explain the positions of some of the jets, it does
not support enough of the observations. The Slow Long Axis Mode fits the
observations quite well, and should be taken to be the correct state for

This work was supported by NASA funds.

15.10-P Ferrante R. F. Moore M. H. Nuth J. A. III
Laboratory Studies of Formaldehyde Polymerization in Cometary
Ice Analogues

Formaldehyde, H(sub)2CO, has been found in the coma of several comets,
and its polymerized form, polyoxymethylene (POM), has been suggested as
a possible constituent in the organic component of Comet Halley. Little
experimental information is available on the processes connecting such
species under appropriate conditions. We have used infrared spectroscopy
to examine the thermal and radiation-induced polymerization of
formaldehyde in mixtures with astrophysically relevant volatile
molecules (H20, CH(sub)30H, etc.) deposited at low temperatures onto low
density sub-micron amorphous silicate grains. Special focus on the
effect of the condensation surface will be attained by comparison with
results employing other substrates (Al, salt, inert surfaces, etc.);
these silicate smokes have previously been shown to promote
crystallization of such ices at temperatures well below those required
with other surfaces. The experiments will progress from simple binary
mixtures, with the ultimate aim of treating multi-component cometary ice
analogues. The presence of Fe, Mg, and other metals in the silicates
will also be examined in the future.

*This work was supported by the NASA Planetary Atmospheres Program
**Supported by the NASA Planetary Materials and Geochemistry Program

15.11-P Verkhoglyadova O. P.
Paraboloidal Model of Formation of Envelopes and Rays in
Cometary Ionosphere

Three-dimension model of large-scale stratification of cometary
ionosphere is studied. Given weak charged particle flows along the tail
in the cometary ionosphere, an aperiodic filamentation instability [1]
results in formation of paraboloidal plasma envelopes in the cometary
head and regular ray structures in the plasma tail. The study is carried
out in terms of linear magnetohydrodynamics in paraboloidal geometry of
the cometary ionosphere [2]. This model enable to study the formation of
both paraboloidal envelopes covering the cometary head and ray system in
the tail. Near the cometary head the instability development gives rise
to formation of a set of paraboloidal envelopes. Moving along the
cometary tail one can obtain the system of quasiparallel rays directed
along the tail axis. Shape, thickness and formation time of these plasma
structures are discussed.

References: [1] Verkhoglyadova O. P., Kotsarenko N. Ya., Pasko V. P.,
Churyumov K. T. (1993) Sov. Astron. Lett., 19(9), 823. [2] Schmidt H. U.
and Wegmann R. (1991) In cometary plasma processes, (A. D. Johnstone
ed.), Geophysical Monograph 61, AGU, 49.

15.12-P Boice D. C. Benkhoff J. Huebner W. F.
A Dusty Coma Model for Long- and Short-Period Comets

The mixing ratios of the minor constituents of frozen gases in the
ice-dust conglomerate of the nucleus is a very important clue to the
original composition of the solar nebula. The usual assumption has been
that the mixing ratios of volatile species observed in the coma is the
same as that in the nucleus when photolytic and gas-phase chemistry is
taken into account. While this may be the case for a solidly compacted
nucleus, it is doubtful for a low density, porous nucleus. Moreover,
evidence from the P/Halley missions and laboratory experiments indicate
that comet nuclei are very porous.

We present multifluid, hydrodynamic si1nulations for the gas, dust, and
plasma flow appropriate for long- and short-period comets. The results
of a sophisticated nucleus surface layer model are used for the initial
gas production rates of the volatile species released from the surface.
The simulations are based on a spherically symmetric neutral coma model
with detailed photolytic and gas-phase chemistry, dust entrainment by
the gas, a separate energy balance for the electrons, separate flow of
the neutral gas, fast neutral atomic and molecular hydrogen, dust
fragmentation, and distributed coma sources of gas-phase species related
to the dust.

These simulations allow a study of how certain coma features, e.g.,
spatial distributions of coma species, neutral and electron
tempera-tures, outflow velocities, etc., change with cometary gas
production and dust-to-gas production rate. In comparison with
observations, the simulations can give insights into the similarities
aud differences of long- and short-period comets.

15.13-P Mueller B. E. A.
A Numerical Study of Complex Spin States of Cometary Nuclei

I investigated published light curve data of comets with respect to
their periodicity with special emphasis on multiple periodicity
analysis. The results were presented at the last DPS meeting [1]. The
conclusion was that we did not have sufficient data for most of the
observed comets to establish with certainty a pure or complex rotation.

For this reason I analyzed numerically generated light curves for comets
in pure and complex spin states using phase dispersion minimization
(pdm) and fourier analysis with clean algorithm (windowCLEAN) methods
[2]. The goal is to find the best strategy to observe the comets in
order to detect complex motion (apart from getting continuous coverage)
and to establish limits on the deviations from simple rotation that can
be detected by windowCLEAN.

15.14-P Schleicher D. G. Sackey S. Ballou R. G.
Spatial Distribution of Cometary Grains

In order to better understand physical and chemical processes taking
place within cometary comae, it is desirable to know the detailed
spatial distribution of both the dust and the various gas species in a
comet. In recent years it has become clear that a significant fraction
of the "dust" coma is composed of material, which is neither inert nor
unchanging. Rather, direct and indirect evidence of the existence of
volatile grains as an additional source of cometary gas comes from the
discoveries of CHON particles and collimated gas jets in Comet Halley.
Analysis of radial brightness profiles of 14 comets by Baum, Kreidl, and
Schleicher (1992) indicated that the majority of profiles were steeper
than the canonical 1/rho, consistent with fading of the grains, perhaps
caused by darkening or shrinking of the grains due to the sublimation of
ices. In order to further investigate this phenomenon, we have an
on-going program at Lowell Observatory of narrow-band CCD imaging of
"bright" comets. In this paper, we will show that Comets
P/Brorsen-Metcalf (1989 X) and Austin (1990 V) both have spatial
profiles consistent with grain fading. Moreover, the deviation from a
1/rho profile for Comet Austin clearly evolves during the month of
imaging. Possible causes of this evolution, such as declining production
rates due to increasing heliocentric distance or effects due to the
change in phase angle, will be presented.

This research was supported by NASA, NSF, and the Research Corporation.

15.15-P Kiselev N. N. Jockers K. Chernova G. P.
CCD Imaging Polarimetry of Comets P/Encke, P/Ashbrook-Jackson,
and P/Schwassmann-Wachmann 2 (P/SW 2)

Using single-aperture polarimetry Chernova et al. (1993) have shown that
the phase dependence of polarization in comets with strong continuum
follows one single curve within measurement error. Comets with reduced
continuum in their spectrum also show reduced polarization and the
values scatter significantly, in particular at large phase angles. In
order to investigate the question if this result is caused by unwanted
contribution of molecules to the measured polarization or if there is
indeed a different kind of dust in the comets with low continuum, we
have performed imaging polarimetry in comets P/Encke (low continuum) and
P Ashbrook-Jackson and P/SW2 (both comets have strong continuum). The
comets were observed at the lm-reflector of Observatory Hoher List.
Germany, and at the 2m-RCC-telescope of the Bulgarian National
Observatory. A red filter (lambda(sub)0 = 694 nm. FWHM = 76 nm) was
used. It includes the faint molecular emissions of NH(sub)2 and H2O^+.

Comet P/Encke was observed at phase angles 51.1 degrees and 80.5
degrees. The degree of polarization found in its sunward fan (10 - 14
degrees) is higher than in other parts of the coma. (4 - 7 degrees), but
less tha.n the value observed in dust-rich comets at similar phase
angles (22 %). In comets P/Ashbrook-Jackson and in P/SW 2 the
polarization is similar to other dust-rich comets. We will discuss the
evidence if the reduced polarization values found in comet P/Encke are
caused by contribution of NH(sub)2 a.nd H20^+ or if the effect is real,
i. e. caused by a different kind of dust grains.

Reference: Chernova. G. P., Kiselev N. N., and Jockers K: 1993,
Polarimetric characteristics of dust particles as observed in 13 comets:
Comparison with asteroids. Icarus 103. 144-158.

15.16-P Keller H. U. Wellnitz D. D. Schulz R. Birch P. V.
A'Hearn M. F.
Gas and Dust Distribution in the Coma of Comet McNaught-Russell

A series of pre-perihelion images of Comet McNaught-Russell was obtained
with the University of Maryland CCD camera system mounted on the 0.61-m
Lowell Telescope at Perth Observatory, Australia. Observations in
several different spectral ranges were taken on 13 nights between March
9th and March 27th, 1994, using standard IHW narrowband filters for CN,
C(sub)2, C(sub)3, blue continuum, and red continuum, and a broad red
bandpass (RG665). The images have been extinction corrected and
relatively calibrated with appropriate standard stars. The light curves
show short-term variabilities in brightness superimposed on the general
increase due to the decreasing heliocentric distance. Surprisingly,
preliminary morphological analysis shows distinct features in the coma
pointing into the sunward hemisphere for CN and C(sub)2, but not for
C(sub)3. Tlle images reflecting the distribution of dust also show no
evidence for distinct features except for the already developed dust
tail. The material distribution as well as the temporal evolution of the
coma in all six components will be presented in a comparative analysis.

15.17-P Osip D. J. Schleicher D. G. Campins H.
The Rotation State of 4015 Wilson-Harrington: Revisiting
Origins for the Near-Earth Asteroids

The rotational characteristics of Near-Earth Asteroids (NEA) have been
found by some to be at least partially diagnostic of their origins.
Binzel et al. (1992) compared the rotation periods of 32 NEAs with those
of the only 5 well obselved cometary nuclei at that time and with those
of their new sample of 32 main-belt asteroids of similar size (d < 5
km). They concluded that the spin rate and shape distributions of the
small main-belt asteroids are statistically indistinguishable from those
of the NEAs while the mean values for the comet sample were
significantly different. Considering their available sample of cometary
nuclei as representative, they concluded that their observations were
consistent with 0-40% of the NEA population being derived from dormant
or extinct cometary nuclei.

CCD photometry perforrned on the comet-asteroid transition object 4015
Wilson-Harrington (W-H) during its most recent apparition have provided
a new rotational light curve and an estimate of the rotational period of
6.1 +- 0.05 hours. The new rotation period and amplitude determined for
W-H are within one standard deviation of the mean values calculated for
the 32 small main belt asteroids by Binzel et al.; the rotation period
is significantly shorter than those of the other comets previously
considered. In addition, Campins et al. (1994) estimate an effective
diameter of 4 km for W-H, similar to those of the NEA and asteroid
samples, whereas the earlier studied comets averaged over 10 km in
diameter. In short, the range in cometary nuclear properties is greater
than indicated by the comets studied previously and although the
statistics are still poor, the size and rotational properties of W-H do
suggest that some fraction of NEAs are of cometary origin.

15.18-P Nazarchuk H. K.
The Photometrical Evolution of Comets

Numerical simulation of the orbital and secular photometric evolution of
cometary nuclei has been carried out using a theory of formation and
destruction of surface dust mantle. As a result of the simulation the
next parameters were obtained: the equivalent radii of nuclei, dust
content, the maximal size of the mineral particles embedded into the
ice. The used model of evolution is rather sensitive to these
parameters. The phorometric curves of eleven comets were calculated for
more than 100 revolutions. All the investigated comets belonged to one
of the next three types: the comet with the perihelion distance q>2 AU
and low dust contents, the comet with q~1.5 AU and ~5% dust material,
the comets with q~0.5 AU, the dust contents ~10% and large nuclei ~10
km. Each of these types has its own features of evolutions.

15.19-P Shulman L. M.
Effect of the Solar Spectrum on Spectra of Comets

If one has a comet spectrum with good enough resolution it
is easy to take a gap between two deep neighbor solar
absorption lines as an "emission" comet feature. The genuine
physical characteristics of comet is not the distribution of
energy in its spectrum but rather the dependence of the
optical thickness of its atmosphere on the wave length. This
function can be obtained by division of the comet spectrum
by the spectrum of the Sun reduced to the same resolution.
this procedure being applied to Comet Halley gave deficiency
of the equivalent widths of the Fraunhofer's lines scattered
by the cometary atmosphere. The luminescence of CHON
particles has been proposed to explain this phenomenon.
There is a suspicion (M. Mumma, private communication) that
this effect may be caused by underresolution of the comet
spectra. This problem is analyzed and discussed.

15.20-P Chorny G. F. Nazarchuk G. K.
Luminescence of Organic Molecules in the Halley Comet

It has been previously observed (December 1985, G. K.
Nazarchuk) that there are two broad features with the
centers at 3950 angstrom (-10, +40 angstrom) and 5100
angstrom (+/- 50 angstrom) in the continuum of the Halley
comet. They are parts of the scattered solar continuum. The
feature at 5100 angstrom has been observed as far as 2000 km
from the nucleus. The feature at 3950 angstrom was more
longlived and has been observed at the nucleocentric
distances up to 20000 km. The assumption is that there are
luminescence features of unknown origin. The luminescent
spectra of some alternative sources in the optical regions
are qualitatively analyzing: (a) the sputtering of water icy
particles at temperature below 100 K by electrons with the
energy between 20 and 300 eV; (b) continuous spectrum of
C(sub)2H radical that arises under photolysis of the
C(sub)2H(sub)2 molecules; (c) the photoluminescence of
lonsdeleite crystals, which are the white grains of diamond
and can be included in composition of graphite particles;
(d) the photoluminescence of the polycyclic aromatic
hydrocarbons (PAHs). The analysis shows that the possible
contribution of (b) and (c) sources is sufficiently slight.
Process (a) can contribute to the detail of 3950 angstrom if
there are the regions in a cometary atmosphere in which the
temperature of the icy particles is in the range from about
77 K to 100 K. The photoluminescence of the small PAHs, in
environment where these PAHs are slightly pushed by H2O and
OH molecules, is the most probable. The candidates among
PAHs are anthracene (C14-H10), pyrene (C16-H10), anthrol
(C14-H9-OH) and other small PAHs with adjacent O, OH, or

15.21-P Shvidkiy A. A. Kotsarenko N. Ya. Lizunov G. V.
On the Theory of Generation of Electromagnetic Waves on
Half-harmonics of Electron Cyclotron Frequency in Space Plasma

Electrostatic plasma waves at half-harmonics of cyclotron frequency are
of special interest. This kind of emission is observed in the Earth and
planetary magnetospheres, including the Jupiter magnetosphere, some
spectra of the Solar radiobursts, laboratory and active space plasma
experiments, etc. From the theoretical point we have to do with a
definite frequency spectrum of Bernstein modes. The problem is to
explain its character, which is common for dissimilar plasmas.
Particularly, the location of spectrum maximums near the (eta +
l/2)omega(sub)c values (omega(sub)c is electron cyclotron frequency, eta
= 1,2,3,...) is of interest. Earlier, a variety of mechanisms were
invoked for this purpose. Most of them deal with the instabilities of
special kinds of plasma distribution functions, like a loss cone,
oscillatory, or temperature anisotropy functions. These approaches could
explain isolated data, but not the general nature of phenomena. The
purpose of this report is to show that a wave generation at frequencies
omega = (eta + 1/2)omega(sub)c is.caused by interaction of electron
flows with Bernstein modes. The linear theory of instability and
estimation for the level of quasilinear instability saturation are
developed. The essence is that the spectrum shape is determined by
dispersion properties of these modes in themselves. An electron flow
plays a role of the energy source only, providing the necessary sign of
plasma wave dissipation.

15.22-P Churyumov K. I. Kleshchonok V. V.
Spectral observations of Comets Scorichenko-George (1990 VI),
Levi (1990 XX), Tanaka-Machholz (1991X), Swift-Tuttle (1992t)
and Shaumasse (1992x) with the 6-M Reflector BTA

Five comets Scorichenko George (1990 Vl), Levi (1990 XX),
Tanaka-Machholz (1992 X), Swift-Tuttle (1992t) and P/Shaumasse (1992x)
were observed with the 6-m renector BTA of Special Astrophysical
Observatory of RAS in Nyzhny Arkhyz during 1990-1993. In the spectra of
comet Scorichenko-George (1990 VI) obtained on March 26/27, 1990,
emission bands of triplet and Asundi of neutral CO were found and
emissions of the negative ion of the carbon C2- were discovered. In the
spectra of comets Levi (1990 XX) (on March 23-24, 1990) and
Tanaka-Machholz (1992t) (on April 29/30, 1992) typical cometary
emissions (CN, C2 et al.) were identified. In the spectra of comet
P/Schaumasse (1992x) (on March 14-15, 1993) anomaly relative gas
productivity of the molecule of C3 (vs. molecule of CN) was found.
Peculiarities of spectra of five comet are discussed.

15.23-T Xie X. Mumma M. J. Bockelee-Morvan D. Colom P. Crovisier J.
Gerard E.
Saturation of the OH 18-cm Lines in Comet Levy

The optical depth of the 1667 MHz transition was found to be
nonnegligible for comet Levy in Sept 02-03,1990, which leads to about a
25% reduction of its gas production rate from the initial value. The
Monte Carlo method is used in simulating the observed line profile
(Bockelee-Morvan et al. 1991), which was fitted with an angular
distribution. The line simulation is based on the OH kinematic
parameters obtained from a full Monte Carlo Particle Transport model
(Xie, 1994) and incorporates the optical depth effect of the 18-cm lines
in an expanding coma.

SESSION 16 ....... Mars Atmosphere I
Wednesday, 8:30 - 10:00 Crystal Ballroom B
R. T. Clancy and R. E. Johnson, Moderators

16.01 James P. B.* Clancy R. T. Lee S. W. Martin L. J.
Monitoring Mars with the Hubble Space Telescope: 1994

Mars may be observed by the Hubble Space Telescope after August 6, 1994
when its elongation exceeds 50 degrees. A regular program of monitoring
of the planet will begin at that time; Mars will be observed every 15
degrees of L(sub) using filters at 255, 336, 410, 502, and 673 nm.
These observations will continue a program of HST observations of Mars
which started in December, 1990. Comparison of our first 1994
observations (L(sub)s = 335 degrees) with the earliest WFPC images will
be used to investigate the occurence of a major dust storm in the
preceding summer reported by Clancey (private communication).
Preliminary results of this comparison will be presented.

This work was supported by STScI Grant GO-5493.

16.02 Rosenqvist J.* Titov D.
Water Vapor Spatial Variations on Mars Inferred from ISM/Phobos
2 Observations

The infared spectrometer aboard the Phobos 2 spacecraft provide a data
bank constituted by 40,000 spectra of Mars in 1989 with a typical
spatial resolution of 25 km. Although the spectral power was modest
(60), the spectral range (0.76-3.15 microns) associated to a very high
S/B (500-1000) pemits to detect H2O at five wavelengths (1.14, 1.38,
1.86, 2.55 and 3.0 microns). Different problems impose to use only two
of them (1.86-2.55 microns). Because these bands could be contaminated
by the surface, we should be very careful in interpreting these
features. The correlations with the altitude of the surface and between
the two bands are nevertheless a strong argument in favor of a
quasi-pure atmosheric origin of these spectral features. We will discuss
the results concerning the spatial variations of water vapor from place
to place as a function of the albedo and the surface elevation and

16.03 Titov D. V.* Rosenqvist J.
Regolith Atmosphere Water Vapor Exchange on Mars: Evidences
from the ISM (Phobos) Measurements and Theoretical Analysis

The measurements of the infrared mapping spectrometer ISM ( Phobos2
orbiter provided valuable data on the atmospheric content on Mars, its
spatial and local time variations /1/. Some peculiarities of vapour
distribution were observed: l)a facto higher H20 mixing ratio above
Tharsis volcanoes as compare surrounding plateau /2/; 2)a factor of 2+ /
- 1 increase of the vapour column density from morning towards noon in
some regions /1/; 3)positive correlation of the atmospheric water amount
with the surface albedo and negative-with the thermal inertia /1/. These
features were interpreted as a result of regolith-atmosphere vapour
exchange caused by the diurnal surface temperature variations and
differences in properties.

A theoretical analysis of the diffusion equation led to the solution for
the rate of water release to the atmosphere. Surface temperature and
regolith buffering coefficient were found to be the main parameters,
which control this process. Montmorillonite clay and basalt grains were
found considered as possible analogs of the Martian regolith. The
difference in their adsorbing properties results in significant
variations of diffusion time scale: the response of the clay regolith to
the diurnal temperature changes is much faster than that of basalt soil.
The models, proposed on this basis, succeeded to explain observed
peculiarities of the water vapour distribution on Mars.

REFERENCES: l.J.Rosenqvist et al. Icarus, 98, 254-270, 1992. 2.D.V.Titov
et al. Submitted to Planet. Space Sci., 1994.

16.04 Krasnopolsky V. A.* Bjoraker G. L. Mumma M. J. Jennings D. E.
High-Resolution Spectroscopy of Mars at 3.7 and 8 Micrometers:
D/H Ratio and Upper Limits to H2O2, CH4, H2CO and HCl

A combination of the KPNO 4 m telescope and the Fourier Transform
Spectrometer with the GSFC postdisperser was used to observe
low-latitude regions of Mars in spectral ranges of 2650-2800 cm^-1 and
1229-1237 cm^-1 with resolving power nu/delta nu = 2.7 x 10^5 and 1.2 x
10^5, respectively. The main spectral features are isotopic lines of
CO2, which make possible determination of the ^18O/^16O, ^17O/^16O and
^13C/^12C ratios on Mars. The,observed HDO lines result in the HDO
abundance which being compared with that of H20 measured by the Vikings
at the same season and latitudes yields an enrichment in the D/H ratio
of a factor of 5.5 + 1 relative to the Earth. Substantial improvement of
upper limits to mixing ratios of some species has been achieved: H2O2 <
35 ppb, CH4 < 170 ppb, H2CO < 5 ppb, HCl < 2 ppb. The H2O2 limit imposes
some constraint to models of the martian photochemistry, while the other
ones preclude any significant chlorine and CH4-H2CO atmospheric
chemistries. Telluric absorptions correspond to abundances of H20 of 2
precipitable cm, CH4 of 2.0 ppm, N20 of 60 ppb, H2CO < 0.6 ppb, HCl <
0.4 ppb. All upper limits are given for the 2 sigma uncertainty.

This work was supported by RTOP 196-41-54 under the NASA Planetary
Astronomy Program.

16.05 Johnson R. E.* Pospieszalska M. K.
Calculations of Pick-up Ion Sputtering of the Martian

The importance of pick-up ions formed from escaping O atoms in the solar
wind interaction region has been pointed out in a number of recent
papers^1,2. Here we re-evaluate our initial estimates ^3,4 of the number
of atoms ejected from the atmosphere per O+ ion incident onto the
atmosphere. This is done using realistic cross sections in a Monte Carlo
particle tracking calculation^4. Results will be shown for an atmosphere
dominated by atomic 0. These are self-consistent calculations in that
the expansion of the Martian corona due to the incident flux of ions is
included in the calculation as well as the escape of 0. We will use
these results to re-evaluate the pick-up ion impact flux and the direct
loss of pick-up ions. That is, the sputter-expanded corona and enhanced
escape flux act to stand-off the solar fields at larger distances from
Mars, and the ion formation rate depends on the O densities in the
ionization region. Therefore, the pick-up ion bombardment rate is also
determined by the atmospheric sputtering yield^4 when sputter escape
dominates other loss processes, as is the case in the early epochs^1.
Evaluating this feed-back process will allow improved estimates of the
atmospheric loss rate.

REFERENCES: [l] J.Luhmann, R.Johnson, M.Zhang GRL 19, 2151, 1992 [2]
B.Jalcosky, R.Pepin, R.Johnson, J.Fox, Icarus in press, 1994 [3]
R.Johnson, JGR 97, 13911, 1992 [4] R.Johnson Space Sci.Rev., in press,

16.06 Kass D. M.* Yung Y. L.
Atmospheric Loss from Mars Due to Solar Wind Induced Sputtering

Since Mars does not have a strong intrinsic magnetic field, the
atmosphere is eroded by solar wind induced sputtering. In this process,
ions (especially O+) reimpact the upper atmosphere after being
accelerated by the interaction of the solar wind and interplanetary
magnetic field. In the process, they acquire significant amounts of
energy (upwards of 1 keV) and will collisionally eject neutral elements
of the atmosphere. Luhmann et al. (1991) showed that early solar system
conditions, especially the much greater EUV solar flux and stronger
solar wind, greatly increased the precipitating ion flux. They used a
sputtering model to estimate that ~ 0.14 bar of CO2 and ~ 50 m of water
could be lost over the last 3.5 Gyr of Martian history.

We use a general Monte-Carlo type atmospheric sputtering model adapted
to Mars to calculate the sputtering efflciencies. Excluding polyatomic
dissociation, the model uses elastic collisions with anisotropic
scattering functions. The model includes two major factors that the
analytic model of Luhmann et al. appears,to have neglected. Their model
appears to ignore some of the secondary collisions that allow particles
to escape. It also treats C atoms as only having their own cross-section
and not that of the full CO2 molecule. With our new sputtering
efflciencies, we calculate that ~2.3 bars of CO2 and ~60 m of water have
been lost over the last 3.5 billion years.

One explanation for Martian geomorphology, notably the channels, is an
early dense atmosphere to create a greenhouse. Such an atmosphere would
require at least 0.5 bar of CO2 (e.g. McKay and Davis, 1991) and 50 m of
water (Baker et al., 1992). The fate of such an early atmosphere is a
significant problem for the greenhouse explanation. If the thick early
atmosphere existed, the new sputtering losses of 2.3 bars of CO2 and 60
m of H20 could remove such an atmosphere over the history of the planet.
Currently the best constraint is the atmospheric stable isotope data
(Jakosky, 1991). Simple calculation, using just the current atmosphere,
for C^l3 indicate that between 1 and 5 bars of CO2 need to be buried in
the planet during the sputtering to match the isotope measurements.

DMK was supported by an NFS Fellowship and this work was partly
supported by NASA grant NAGW-1538.

16.07 Forget F.* Hourdin F. Talagrand O. Pollack J. B.
Observations and Simulation of the Martian Polar Climate:
Impact of the CO2 Clouds

Analyzing thermal infrared observation from Viking IRTM and Mariner 9
IRIS, we have found strong evidences for the presence of CO2 and H20
clouds above the Martian polar caps during the polar night [1]. This
study revealed that, by altering the thermal radiation emitted to space,
these clouds have a strong impact on the thermal balance of the polar
regions, lowering the amount of CO2 that condense during the fall and
winter seasons. A strong hemispheric asymmetry was observed: in the
northern hemisphere, C02 clouds were very common at most polar latitudes
unlike in the southern hemisphere where they were less frequent below 80
degrees S.

We have tried to take this impact into account in climate modeling with
the Martian Global Circulation Model (GCM) of the Laboratoire de
Meteorologie, Dynamique. For that purpose, we have first developed a
simple scheme to simulate the condensation, advection, and possible
sedimentation or sublimation of the CO2 ice particles in the GCM. This
treatment is based on the calculations of the condensation rates of the
CO2 in the cold polar atmosphere by the GCM and on clouds microphysics
considerations. We have also derived a simple parametrization of the
simulated polar clouds radiative properties from our analyses of the
thermal infrared data.

The results of our simulations can be compared with the Mariner 9 and
Viking observations, and can also be validated by considering the
ability of the model to reproduce the Viking Landers pressure curves,
which mainly depend on its ability to simulate the climate of the polar
region and the associated CO2 condensation/sublimation cycle [2].

References: [1] Forget and Pollack, Bull. of A.A.S., 25-3, p. lO70,
1993. Also, Paper submitted to JGR Planet, 1994. [2] See Hourdin et al.,
J.Atmos.Sci. 50, 3625-3640,1993, Pollack et al., JGR Planet
98(E2),3149-3181, 1993, Talagrand et al., this issue.

16.08 Talagrand O.* Forget F. Hourdin F.
Martian Seasonal and Transient Pressure Variations: GCM
Simulations and Viking Observations

The Martian General Circulation Model (GCM) of Laboratoire de
Meteorologie Dynamique (LMD) has been used to simulate the Martian
Climate and analyze its sensitivity to various parameters (Hourdin et
al., submitted to J. Geophys. Res.). Results of pluriannual simulations
have been evaluated primarily on the basis of comparisons with the
Viking surface pressure measurements.

The model is able to reproduce rather accurately both the large seasonal
variations and the rapid transient oscillations of the surface pressure.
The seasonal variations are due to the change in the the atmospheric
mass (resulting from the condensation of the CO2 atmosphere in the polar
caps), but also to internal latitudinal mass redistributions associated
with the atmospheric circulation (Hourdin et al., J. Atmosph. Sci., 50,
3625-2640, 1993). The simulated internal mass redistribution (found to
be weakly sensitive to model parameters) can be used to retrieve the
seasonal variations of the total atmospheric mass from the Viking
pressure measurements. The periods (typically 2-5 sols) and the
amplitudes of the transient eddies provide a useful diagnostic of the
mid-latitude atmospheric circulation. The model is able to simulate the
decrease of the amplitude of the eddies for increasing optical depths,
as observed near northern winter solstice, particularily during the
1977-B global dust storm. Even without seasonal variations of the
atmospheric dust content, the reduction of the transient activity for
dusty conditions is maximum near northern winter solstice. For large
dust amounts, the increase in the latitudinal extent of the Hadley
circulation strongly reduces condensation at the edge of the forming
cap. But in the same time, the reduction of the transient activity,
responsible for latitudinal energy transport toward higher latitudes,
reinforces strongly condensation at the pole.

16.09 Fanale F. P.* Postawko S. E.
Mass Loading: A Key to the Early Mars Greenhouse?

Carbon dioxide atmospheric greenhouse models for early Mars require
massive CO2 atmospheres and are subject to cloud formation. Similarly,
models augmenting the greenhouse with SO2 molecules run afoul of short
SO2 atmospheric mean residence times (MRT) dictated by the rapid
reaction with OH-. However, the MRT of SO2 is actually a variable which
depends on the relative rate of SO2 supply and H20 resupply. Hesperian
eruption rates can supply SO2 at global rates of 10^-9 to 10^-8 g cm^-2
sec^-l which could destroy the current atmospheric H20 inventory in days
and the higher H20 inventories associated with greenhouse models in less
than 1 month. This, together with recent models allowing wide obliquity
variations, results in a wide range of plausible but unfamiliar models
in which CO2 and H20 can be "traded off" against SO2. We assume than in
early conditions CO2, H20, and SO2 are cold trapped at the same
temperatures. For cap temperatures of 213K and 195K: PCO(sub)2 = 4 bars
and 1 bar; PS0(sub)2 = 50 mb and 12 mb (when equilibrium with the cap is
achieved); atmospheric H20 destruction time = 1 month to 2 days (for the
highest eruption rate); SO2 mixing ratios are on the order of 10^-2. It
is unlikely that the poles or regolith could resupply atmospheric H20 at
a rate comparable to the SO2 supply rate. Indeed, the most viable source
of H20 is the fissure eruption itself. However, it cannot supply the H20
directly and efficiently to the atmosphere as it does the SO2 owing to
the local and regional cold trapping. Steady state between H20 and SO2
can be achieved by creation of an open water lake with a radius of
approximately 200 km or, for an assumed ablation rate of 30 cm/yr, an
ice covered lake of radius 900 km.

SESSION 17 ....... Comets II
Wednesday, 10:30 - 12:00 Crystal Ballroom A
M. S. Hanner and M. H. Moore, Moderators

17.01 Grun E.*
Comet Surface Probes for Rosetta

The Rosetta spacecraft will deploy surface probes on a cometary nucleus
with the goal to study the most primitive material in the solar system.
The objectives of these surface probes are the in situ characterization
of the chemical, mineralogical and physical properties of bulk surface
material and the study of cometary processes at the surface of and
within the nucleus. Four different scenarios of surface probes and
their relationships to the objectives are discussed: (1) A
battery-powered surface station carries a camera for morphologic and
mineralogic studies and several sensors for compositional analysis of
cometary surface material. It transmits pictures and compositional data
for several hours after landing. (2) Two different surface regimes are
known to exist on a cometary nucleus: inactive areas which are depleted
of volatile materials and "active" areas which emit gas and dust at
elevated insolation levels. Either multiple probes or a mobile probe
are required to sample both regimes. (3) Comet material undergoes
continuous thermal modifications due to variable insolation: volatile
components are redistributed and depleted within the nucleus and finally
leave the surface. In order to study these processes a long-term probe
is necessary which both daily and seasonal variations of the physical
and chemical properties of cometary materials. (4) The nucleus is
thermally coupled to the exterior via a thin surface layer. Within an
about meter-thick layer strong temperature variations and both physical
and chemical modifications occur. Studying this heat transfer process
and the corresponding material modifications will require a sub-surface
probe. In order to understand the compositional data obtained at the
surface all of these questions have to be addressed by the Rosetta
Mission. Possible implementations of these objectives will be

17.02 Young E. F.*
Collisional Evolution Within the Kuiper Belt

Various populations of Kuiper Belt objects are studied with a simple
computer simulation. Hill's equations (linearized equations of motion
within a rotating reference frame) are used to describe the objects'
trajectories. Borrowing a trick from Wisdom and Tremaine's treatment of
particles in a ring (AJ 95 no. 3, 1988), this simulation tracks a local
cluster of objects and models the rest of the Kuiper Belt as copies of
this local cluster.

The simulation is a useful tool for studying the following questions:
What is the collision rate in the Kuiper Belt? What is the dust
production rate in the Kuiper Belt, and what is the flux of dust into
the inner solar system?. What initial populations are consistent with
the current assumed Kuiper Belt population? What fraction of Kuiper Belt
objects are the products of collisions? How does the size distribution

Figure 1. appears here in the hard copy.

This work was supported by the National Research Council Associateship

17.03 Capria M. T.* Coradini A. Capaccioni F. Espinasse S.
Orosei R. Salomone M. Federico C.
Differentiation Processes of a Comet Coming from the 'Kuiper

We studied the evolution of a porous comet nucleus injected from the
Kuiper belt to the inner Solar System. The comet undergoes different
successive closeencounters with giant planets, becoming a "Jupiter
family" comet. We have analyzed the differentiation of the nucleus and
the conditions under which a dusty crust forms, during the dynamical
evolution of the comet. Our nucleus is compose of a mixture, initially
homegeneous of different ices and grains. Ices can be considered
amorphous, due to the low temperature of the comet in its formation
area. For the water ice, the transition amorphous-crystalline is
followed in detail. We have used different values of conductivity for
amorphus ice, different gas and dust concentrations, different initial
conditions (presence or not of an original organic crust due to cosmic
ray bombardment) in order to study how this choice affects the thermal
evolution and the differentiation of the nucleus. The gas and dust
emission is studied as well.

17.04 Coradini A.* Capaccioni F. Capria M. T. Orosei R.
Salomone M. Federico C.
Thermal Evolution of a Short Period Comet: Bidimensional Models

The differentiation of comet-like body has been studied for several
years by means of uni-dimension models allowing to solve the heat
equation. These models allowed to study the importance of different
parameters - such as porosity, dust content, presence of different
gaseous components - on the thermal evolution and differentiation of the
cometary nucleus. In this paper we present the thermal evolution of the
comet nucleus obtained solving the hear equation in 2D in a spherical
coordinate system. The nucleus is initially made by amorphous ice and
orbits around the Sun at a distance of 50 A.U. The evolution differs
from the unidimensional one mainly because the latitudinal insulation
differences generate a temperature field in the interior of the nucleus
that is strongly latitude dependent. We study the evolution of the
nucleus taking into account also of the amorphous-crystalline ice

17.05 Malhotra R.*
The Origin of Pluto's Orbit: Implications for the "Kuiper Belt"

Pluto's Neptune-crossing orbit owes its dynamical longevity to a
reasonant libration of its perihelion 90 degrees away from Neptune's
mean longitude. This peculiar orbit plausibly owes its origin to
dynamical processes in the early Solar System. A resonance capture
mechanism is possible during the loss of planetesimal mass from the
vicinity of the giant planets [Malhotra, R. Nature 365:819-21 (1993)].
If this mechanism were in operation during the early history of the
planetary system, the entire region between the orbit of Neptune and
approximately 48 AU would have been swept by first order mean motion
resonances. Thus, resonance capture would occur not only for Pluto, but
quite generally for other trans-Neptunian "planetesimals". Furthermore,
the dynamical erosion of the number density of small bodies in the
"Kuiper Belt" (the putative source reservoir of short-period comets)
would be greater than that suggested by recent studies of test particle
orbital stability [Holman, M. & Wisdom, J. AJ 105:1987-99 (1993);
Levison, H. & Duncan, M., ApJ 406:L35-38(1993)]. Theoretical
predictions for the orbital distribution of trans-Neptunian small bodies
will be presented. A significant implication for observational searches
is that higher number density of objects may be found in very narrow
regions interior to ~ 48 AU, in stable resonant orbits with non-random
distribution of orbital elements.

17.06 Kotsarenko N. Ya.* Rapoport Yu. G. Verkhoglyadova O. P.
Soliton Model of Formation of Condensations in the Cometary
Plasma Tails

Condensations are typical objects in the cometary ionosphere. We assume
that they are stable dense structures caused by nonlinear solitary waves
of alfven and magnetosonic wave branches which play an important role
for space plasma objects. The high level of spectra density of the above
waves in the cometary plasma is in favour of our assumption. In order to
study the possibility of soliton formation mechanism for condensations,
nonlinear dynamics of the magnetosonic waves is considered.

The study is carried out in terms of a set of MHD-equations for ions and
electrons of the cometary plasma. We study the one-dimensional wave
propagation with angle theta to the external magnetic field. The
reductive perturbation method for weakly dispersive and weakly nonlinear
waves [1] is used. The KdV soliton is formed under the defined
conditions for theta and for wave number of the initial magnetosonic
wave. The estimation for solitary wave propagating with theta = 15
degrees in external magnetic field of 10 nT gives the spatial scale of
magnetosonic soliton about ~=750 km which value agrees with the
observational data.

We propose that sperical plasma clouds maybe nonlinear solitary
MHD-waves in the cometary ionsphere.

[1] Kakutani T. (1974) Suppli. Progress of Theor. Phys., N-55, 97

17.07 Farnham T. L.* Meech K. J.
Modelling of Comet P/Tempel 2 Using a Finson-Probstein Fitting

We have developed and are refining a kinetic model of cometary dust
grains, which is based on the Finson-Probstein (F-P) technique, to study
the particle populations in cometary dust tails. The appearance and
morphology of a comet's dust tail are the result of the opposing forces
of gravity and solar radiation pressure that act upon individual grains.
Using the dynamics of the system, the surface density along the tail is
determined by summing up the scattering contributions from all
particles. By matching models to observed images, the various grain
parameters can be determined. The main parameters are the particle size
distribution, the dust production rates and the escape velocities. Our
model is based of the original F-P work, however, several improvements
have been incorporated into our model, including non-spherical emissions
and more realistic scattering functions. The most recent refinement to
the model is the implementation of a fitting process to determine the
parameters which best reproduce the observed images. The fitting process
will be based on the Levenberg-Marquardt method of solving non-linear
least squares problems. It will have the capability of solving for a
number of dust grain parameters, while at the same time determining how
the individual parameters are interrelated in E producing the
morpllology of the tail. This is important in the interpretation of the
grain parameters, because changes in the appearance of the tail produced
1?! by altering one parameter can sometimes be corrected for by
adjusting another. Comet P/Tempel 2 has been chosen as a test case for
the fitting procedures, because it is a periodic comet that has been
well-studied. Its rotation period is known to be 6.5 hours, and the
direction of the rotation axis and the lag angle of the sublimation have
both been calculated, so the orientation of the I? emission can be
well-defined. We have data from 17 different observing runs a between
March 1987 and March 1992 (spanning one full orbit). The full orbital
coverage allows the development and dissipation of the tail to be
observed as a Y fan shaped tail has formed by April 1988, and remains
well defined until January 1989, when it begins to dissipate. There is
excellent coverage of the comet D near its perihelion passage, which is
the best time to constrain the properties A of the particles. The
application of suitability tests indicates that P/Tempel 2 A is a good
candidate for modelling using our technique, especially with the large
time base that the images cover, It will make a good test candidate for
the S fitting procedures because so much is already known about the
properties of the nucleus. A review of the modelling process, with the
details of the fitting D procedures will be presented, along with the
results of the application of the S model to the observations of Comet
P/Tempel 2.

17.08 Benkhoff J.* Huebner W. F.
Influence of the Gas Flux on Temperature and Density in the
Near Surface Layers of Comets

We calculate the gas flux of the volatile icy components in the surface
layer of a porous, short-period, Jupiter-class comet, in order to
investigate the relationship of the observed relative molecular
abundances in the coma with those in the nucleus. The model assumes a
body containing dust, one major ice component (H20), and two minor
components of higher volatility (C0 and CO2). The mass and energy
equations for the different volatiles are solved simultaneously with
appropriate boundary conditions. The model includes inward and outward
flowing gas within the body, escape of outward flowing gas from the
body, depletion of less volatile ices in outer layers, recondensation of
gas in deeper, cooler layers, and changing porosity as a result of
sublimation. The dust layer at the surface of the nucleus drastically
reduces the sublimation flux of H20 at heliocentric distances less than
about 3AU and keeps the the mixing ratio of the gas flux of minor
volatiles to that of H2O from the nucleus into the coma allmost
constant. In contrast, the mixing ratio of a dust free comet varies by
several orders of magnitude throughout its orbit. Thus, the relative
abundances of species observed in the coma are in most cases not the
same as those in the nucleus.

We also calculate temperature and abundance distributions in the surface
layers of the nucleus as a function of diurnal variation of the incoming
solar flux for various positions in the orbit. Diurnal deviations of
temperature observed up to a depth of about 2m assuming a thermal
conductivity of about 2W/mK. For a thermal conductivity of about
0.02W/mK the penetration depth of the diurnal temperature wave is less
than 0.2m. Results of abundance distributions indicate that it will be
impossible to determine the relative abundances of ices more volatile
than water from samples taken a few meters below the surface during a
comet rendevouz mission, because the surface layers are depleted of the
minor volatiles.

We gratefully acknowledge support from NASA Planetary Atmospheres
Program grant No. NAGW-2205 and DFG grant No. Be1630/1-1.

17.09 Combi M. R.*
Direct Monte Carlo Simulation for Tenuous Planetary
Atmospheres: 1-D and 2D Cometary Coma Outflow

The method of Direct Monte Carlo (DMC) simulation has been applied to,
and specific adaptations developed for, a general class of tenuous
planetary atmosphere problems. The motivation of the method begins with
the same assumptions that are appropriate in the general and fully
kinetic form of the Boltzmann equation for dilute gases (i.e., binary
collisions), from which the usual time-dependent hydrodynamic equations
(Euler and Navier-Stokes) can also be denved in the high density limit.
By employing the DMC method, restnctive assumptions such as the form of
the molecular phase space distribution function need not be made. DMC
simulation affords a method for studying time-dependent,
multidimensional non-equilibrium gas-flows with arbitrary molecular
distribution functions. Applications will include cometary coma outflow
as well as gas escape from gravitational bodies such as Io. The method
has a long history in subjects as diverse as shock structures in
aerodynamic flows, star cluster evolution, and the earth's ionosphere.
The DMC simulation is dishnguished from past atmospheric Monte Carlo
coma and exosphere models which are generally of the "test-particle"
type, whether they are called particle trajectory models, hybrid
gas-dynamic/Monte Carlo, or full Monte Carlo. Methods for treating
boundary conditions, binary collisions, photochemistry, chemistry,
internal degrees of freedom, gravity, as well as the use of weighted
particles for widely disparate species densities have been addressed.
The model has been verified by comparison with l-D spherical and 2-D jet
gas-dynamic expansion calculations in the high-density limit, and with
1-D hybrid gas-dynamic/Monte Carlo calculations in the case of realistic
bright-comet densities. In this introduction to the subject, the
applications to one-dimensional spherical and two-dimensional
axisymmetric, steady-state, five-species (OH, H(sub)2, 0 and H) water
comae are described, and the breakdown from the hydrodynamic assumption
of local equilibrium is explored.

17.10-T Moreels G. Rousselot P. Goidet-Devel B. Clairemidi J.
An Extended Source of Small Grains in C2 Molecules in Halley's

The existence of an extended source in the coma of Halley's comet is
shown in a straightforward manner by two independent methods. The first
one consists to calculate the ratio of the solar dust-scattered
intensity at two wavelengths: 377 and 482 nm. Using the spectra of the
Vega2 three-channel spectrometer, an image of the scanned field of view
shows an excess of near UV intensity in a diffuse region located at a
cometocentric distance of 20000 km. A model based on Mie scattering
shows that the fragmentation of dust particles producing submicronic
grains may result in a slight increase of the 377/482 intensity ratio.
The second method consists to visualize the (1,1)l(0,0) intensity ratio
of the two most intense bands of the delta v = 0 sequence of carbon at
513-516 nm. The image of this ratio remarkably well reproduces the
features of the image showing the color of dust. The interpretation is
the following. Dust particles are shown to be transported in collimated
jets to cometocentric distances longer than 40000 km. During this time,
they suffer sputtering and fragmentation processes and release different
gases: OH, NH, CN and C2 that form gaseous jets. The observed extended
source around 20000 km is depicted by a high local production rate of
submicronic grains and release of freshly produced C2 radical having an
initial high rotational temperature.

SESSION 18 ....... Mars Atmosphere II
Wednesday, 10:30 - 12:00 Crystal Ballroom B
J. R. Murphy and W. W. McMillan, Moderators

18.01 Harri A.-M.* Pirjola R. Pellinen R.
Meteorological Research by Microsensors on the Surface of Mars

There are three main reasons which make observations of meteorological
phenomena in the Martian atmosphere important: - Meteorology as science
including comparative meteorology (Mars - Earth). - Other instruments
need information about meteorological parameters. - Weather conditions
essentially affect planning of safe future landings on Mars.

The Earth and Mars have almost equal rotation periods and inclination
angles of the axis, and hence similarities in the meteorology of these
planets are considerable. However, the atmosphere of Mars is very thin
and solar irradiation is low compared to that of the Earth. It means
that Mars constitutes a simple "meteorological laboratory". At present,
the number of theoretical results of Martian meteorology is much larger
than measured data, so measurements are urgently needed. Both in-situ
recordings on the planetary surface and remote sensing from an orbiter
are necessary. The former give important ground-truth calibrations to
the latter.

This paper concentrates on meteorological instrumentation for Mars using
capacitive microsensors developed by the Vaisala Ltd, Finland. Due to
their accuracy, light weight and low power consumption, they are very
suitable for planetary exploration purposes. They are also manufactured
in quantities enabling a procurement of a large number of sensors, which
can then be qualified and screened to meet the required criteria.

Vaisala sensors will be used in pressure, humidity and temperature
measurements in the Mars'96 mission (former Mars'94), and plans of their
use in forthcoming missions exist as well. Experiences obtained at the
Finnish Meteorological Institute on Martian meteorological
instrumentation including their testings and calibrations are described
in this paper, and some suggestions for future missions are also given.

18.02 Martin L. J.*
Martian Clouds Without Mountains

In the new Martian nomenclature, "fossae" is the AP designation for
areas with numerous grooves, trenches, or ditches, as seen on Mariner 9
and Viking imaging. These have little or no relationship to the linear
albedo features (real and imagined) that Lowell and others called
"canals". Fossae are generally not set off by their albedo and do not
have enough relief to be identified from Earth. Most fossae are also
much shorter than Lowell's canals, although some around Alba Patera
stretch to over 1000 Km.

Many of the areas on Mars that have recurring clouds have often been
associated with large volcanoes, primarily the "W" clouds in the Tharsis
area. We had been troubled, however, that the Alba Patera area, which is
a very low relief volcano, often was covered by more prominent clouds
than the large, high mountains of Tharsis. We were also aware that some
of strongest clouds in the "W" formation were not over mountains of any

In re-examining cloud studies of International Planetary Patrol data
from 1969, we looked at other areas where clouds appeared most
frequently. It was found that nearly all of these areas where clouds
reocurred throughout the study period, were areas whose topography
included fossae. Furthermore, in going back to the Viking imaging, we
were reminded that the most discrete clouds at the Tharsis Montes appear
to have originated from fossae on the slopes of, or near these
mountains. We suspect that these great edifices may have little to do
with cloud (orographic) formation, but rather that they are also the
locations of the fossae which actually spawn the clouds.

18.03 Schmunk R. B.* Allison M.
Potential Vorticity Profiles of the Mars Atmosphere

Potential vorticity (PV) maps are becoming increasingly important
diagnostic tools in the study of Earth's atmospheric circulation, but
the usefulness of their extension to Mars is as yet unclear due to its
strongly diabatic circulation. We present results of numerical
simulations by the Goddard Institute Mars general circulation model
(Allison et al. 1993, in LPI Tech Report 93-05), paying particular
attention to meridional profiles of the Ertel potential vorticity. At
altitudes below the 0.5-mbar level, the PV is seen to remain at values
near zero, but at altitudes above that level, the PV changes rapidly,
establishing what has been termed a dynamic tropopause. The largest
anomaly observed in various seasonal profiles in the PV profiles occurs
near 60 degrees N at L(sub)s = 270 degrees (onset of southern summer)
and corresponds to a strong westerly jet. The possible utility of PV
inversion as a tool for the diagnosis of thermal sounding data will also
be discussed in light of the GCM simulations.

* NRC Postdoctoral Research Associate.

18.04 Barnes J. R.* Haberle R. M. Pollack J. B.
Low-Latitude Circulations in the Mars Atmosphere as Simulated
by the NASA-Ames GCM

It has long been known that much of the dust storm activity on Mars
originates in low latitudes -- in the tropics and subtropics. This
region always contains the sub-solar latitude, so the direct thermal
forcing needed to drive such storms is at a maximum. Modeling studies
have shown that the Hadley circulation on Mars exhibits its strongest
low-level winds in low-latitudes, and the thermal tides tend to have
large near-surface amplitudes in the highlands of the southern tropics
and subtropics. The intensity of both of these circulations is strongly
dependent upon the atmospheric dust loading. Recent analyses of
numerical simulations performed with the NASA-Ames Mars GCM indicate the
probable importance of several other aspects of the low latitude
circulation in the development of dust storm activity. One of these is
the existence of regional intensifications in the lower branch of the
Hadley circulation, associated with the very large topography in low
latitudes. These have been referred to as "western boundary currents" in
several recent studies with simplified models. In the GCM simulations
they indeed appear to have some of the basic characteristics of such
flows, but they are modified somewhat by thermal influences. These
"boundary currents" tend to be strongest on the eastern flanks of the
Tharsis region, and are found to increase considerably in strength with
increasing dust loading in the GCM simulations. The boundary current
winds are strongest near northern winter solstice, as is the Hadley
circulation itself. Since these wind systems are low-level in nature it
might be expected that they show significant diurnal variations, and
these are clearly evident in the GCM results. At least in some regions,
slope winds appear to be a major factor in the diurnal variability of
the boundary currents.

A second important aspect of the low-latitude circulations in the GCM is
the intensity of a subtropical westerly low-level jet, especially during
summer in the southern hemisphere. The existence of this jet and its
association with the Hadley cell have been noted in previous studies; it
is much stronger in southern summer than in northern summer. While this
difference is basically tied to the increased vigor of the Hadley flow
at southern summer solstice, topographic asymmetries between the
northern and southem subtropics (which contribute strongly to the
differences in the Hadley cell) play a role in the regional structure of
the jet. The time mean zonal winds in this jet reach speeds of over 50
m/s at the lowest GCM level (250 m) in southern summer; these speeds
increase by a relatively small amount with dust loading. When tidal and
slope winds are "superimposed" on this jet, regional wind speed maxima
as high as 80-90 m/s are found at the lowest GCM level during southern
summer under dusty conditions.

18.05 Siili T. T.* Haberle R. M.
On the Latitude Dependence of Martian Slope Winds

One of the factors affecting the circulation of the Martian atmosphere
is the topography of the planet. The surface relief is pronounced and
large expanses of sloping terrain occur. At such terrains the surface
heating and cooling drive up- and downslope flows, respectively, through
buoyancy changes in the airmass above the surface. These slope winds
tend to occur at areas and seasons where in addition to the presence of
slopes large diumal temperature variations occur and weak ambient winds

The thin C02 atmosphere (during low-dust conditions) absorbs little of
the incoming solar radiation and is hence strongly controlled by the
surface -- leading to e.g. large diurnal temperature variations. During
the Martian summer the large-scale winds are also relatively weak.

At least summertime slope winds can thus be expected and indications of
regular slope winds were indeed observed by the Viking Landers [Hess at
al., 1977]. Relevant modelling works have been e.g. Blumsack [1973], Ye
et al. [1990], and Savijarvi and Siili [1993].

The latitude affects the slope wind characteristics through the Coriolis
force and at high latitudes also through the modification of diumal
heating and cooling characteristics. The works of Savijarvi and Siili
[1993] and Haberle [1993] on latitude dependence of the slope winds have
indicated modest to strong amplification at around 30 degrees N and near
equator, respectively.

The goals of this work have been on the one hand to identify sites,
where the slope winds are strong,enough to contribute significantly to
dust raising or where high winds may affect mission landings (e.g. Mars
Pathfinder), on the other hand to test some assumptions and
parametnsations inherently necessary in l-D but not in 2-D models. The
high-resolution 2-D mesoscale model has also allowed a more focussed
study of topographical fearlres than the the large gnd lengths of GCM's

Work of T. Siili (permanent affiliation Finnish Meteorological
Institute) is supported by an ESA Extemal Fellowship, BLUMSACK, S.L. et
al., 1973, J. Atmos. Sci., 33, p. 66-82, HABERLE R.M., 1993: Unpublished
l-D model results, HESS S.L. et al., 1977, J. Geophys. Res. 82:28, p.
4559-4574, SAVIJARVI, H. and SIILI, T., 1993. J. Atmos. Sci., 50, p.
77-88. YE, Z.J. et al., 1990. J. Atmos. Sci., 47, p. 612-628.

18.06 Greeley R.* Leach R. N. Lacchia B. M. White B. R.
Trilling D. E. Pollack J. B.
New Estimates of Minimum Wind Speeds for Raising Dust on Mars

Dust storms have been observed on Mars for decades. Previous estimates
for threshold wind speeds were ~150 m/s for 1-2 micrometer dust grains;
yet, strongest winds measured by the Viking landers were only ~40-50
m/s. Moreover, GCM maximum predicted winds are ~70 m/s during summer in
the southern hemisphere. To address this apparent discrepancy, wind
tunnel experiments were conducted in martian atmospheric densities using
1-2 micrometer-size dust. One experiment series involved an
aerodynamically smooth test bed. The second series involved a test bed
covered with 1 cm rocks spaced 1-2 cm apart to create a rough surface.
Even at the maximum wind tunnel speed of 145-m/s, dust was not suspended
from the smooth floor; rather, the dust rolled into small balls and
cylinders to the end of the test bed. In contrast, wind speeds as low as
~30 m/s (u(sub)* = 2.2 m/s) were able to lift dust from the tops of
pebbles on the rough test bed. Some material blew off the rock tops and
collected between the rocks or adhered to downwind faces of rocks.
Nonetheless, estimates of the total flux of material lifted into
suspension from the rocky surface is sufficient to-account for observed
dust storms on Mars. Consequently, we suggest that dust storms are
likely to be generated on Mars from any dusty, rocky surface where
minimum wind speeds of ~30 m/s occur, and that martian dust storms can
be explained without resorting to ad hoc mechanisms, such as those
involving dust devils or devolatilization from the subsurface. This work
was supported by NASA-Planetary Geoscience.

18.07 Murphy J. R.* Bridger A. F. C. Pollack J. B.
The Modal Response of the Martian Semidiurnal Thermal Tide
During Dust Storms

The semidiurnal thermal tidal response of the martian atmosphere is
strongly coupled to the quantity and spatial distribution of suspended
dust. Temporal variations in the observed Viking lander surface pressure
semidiurnal amplitudes and latitudinal variations of the same are
indicators of the rapidly changing atmospheric dust load during global
dust storms.

The semidiumal tide is represented zonally by wavenumber two and
meridionally by an infinite number of modes (meirdidional structures)
which take the numerical form of Hough functions, which are themselves
the weighted sums of normalized Legendre polynomials. We have performed
a Hough function decomposition of the surface pressure and atmospheric
heating rate fields calculated in a series of numerically simulated
Martian global dust storms to ascertain the modal structure of the tidal
response and its temporal variation. These decomposed fields are
directly related to the spatial distribution of the atmospheric heating
generated by the suspended dust, which is the tide's forcing mechanism.

The calculated modal responses indicate that equatorally symmetnc modes
are preferentially enhanced but that the lowest order (smoothest)
equatorially asymmetnc mode (the [2,3] mode) also increases in amplitude
during the initial sols of the simulated storms when dust is still
preferentially contained within the southern (dust source hemisphere. We
will show how these modal results produce northern hemisphere surface
pressure tidal signatures like those observed by the Viking landers, and
how a network of pressure sensors can be employed to define the tidal
modal structure and by inference the spatial distnbution of suspended

This work is supported by NASA's Planetary Atmosphere's Program,
RTOP-154 20-80-16 and RTOP- 154-30-80-16.

18.08 Clancy R. T.* Lellouch E. Billawala Y. N. Sandor B. J.
Rudy D. J.
Microwave Observations of a 1994 Mars Global Dust Storm

Microwave CO spectra of Mars, obtained at the Kitt Peak NRAO telescope
in Arizona and the IRAM telescope in Spain in the spring of 1994, reveal
20-30 K warming of the globally averaged Mars atmosphere over the 0-50
km altitude range. This warming is similar to that observed by the
Mariner 9 and Viking 15 micrometers CO2 measurements of temperature at
~25 km altitude in association with the 1971 and 1977 Mars global dust
storms, respectively. The January-June 1994 microwave observations
consist of whole-disk 1.3 and 2.6 mm spectral line measurements of the
pressure broadened, LTE emission/absorption by Mars atmospheric CO.
Analyses of such spectra, assuming that the photochemically long-lived
CO molecule is well mixed in the Mars atmosphere, allow retrievals of
globally averaged Mars atmosphenc temperature profiles over the 0-50 km
altitude range with roughly 10-15 km vertical resolution (e.g.; Clancy
et al., L Geophys. Res., 95, 14543-14554, 1990). We observed a very
dramatic reduction in the CO 1.3 mm spectral line depth and width
between January 14 and Apnl 8-9 of 1994 at Kitt Peak. This time frame
corresponds to a change in Mars solar longitude (L(sub)s) from 204
degrees to 254 degrees, consistent with most observations regarding the
primary Mars season (southern spnng/summer) for initiation of Mars
global dust storms (e.g.; Martin and Zurek, J. Geophys. Res., 98,
3221-3246, 1993). Additional measurements at IRAM for both the 1.3 and
2.6 mm CO lines and at Kitt Peak for the 2.6 mm line on May 10-11
(L(sub)S = 275 degrees) indicate that the large atmospheric dust heating
persists for altitudes below 40 km, but that the 40-50 km region may be
cooling. This suggests that the dust loading may be decreasing at the
higher altitude levels by L(sub)S = 275 degrees. Continuation of these
CO spectral line observations, including IRAM spectra obtained on May 24
and June 8, should provide a unique definition of the thermal evolution
of the Mars atmosphere over the 0-50 km altitude region in response to a
global Mars dust storm. These microwave observations will be
supplemented by Hubble Space Telescope imaging and spectroscopy of Mars,
to begin in September of 1994 (L(sub)S~ 340 degrees) (see James et al.,
Icarus, in press, 1994).

18.09 Ockert-Bell M. E.* Pollack J. B. Bell J. F. III
Wavelength Dependence of the Radiative Properties of Martian
Atmospheric Dust

This research investigates the importance of the wavelength dependence
of radiative properties such as the imaginary index of refraction and
single-scattering albedo throughout the visible and near-infrared
wavelengths. A solar average value for these quantites may not provide
an accurate representation of dust light scattering and absorption
because of potentially large changes in these quantities at solar
wavelengths. Knowledge of the wavelength dependence of light absorption
and scattering properties may provide a better definition of the
influence that the atmospheric dust has on heating of the atmosphere.

We have analyzed visible and near-infrared composite spectra from
Mustard and Bell [1]. The data set, taken in 1988 and 1989, covers a
wide range of wavelengths (0.4 to 3.0 microns). The spectra were
augmented to include data from 0.2 to 0.4 microns [2] and 3.0 to 4.0
microns [3]. For this investigation we operate under the assumption that
the properties of the atmospheric dust closely mimic those of the
"bright" soil on the surface. Since the optical depth of the atmospheric
dust was low during the time period of the data acquisition, we can use
Hapke theory [4,5] to extract the single-scattering albedo of the soil.
By scaling the imaginary index of the soil to agree with the estimates
of the atmospheric dust in the visible (0.55 microns) [6], we derive the
spectral dependence of the real and imaginary index of refraction and
the single-scattering albedo in the entire visible and near-infrared

The results of this inquiry will be presented. We have found that there
is a strong spectral variation for the imaginary index and albedo of the
dust over the wavelengths that we have investigated (visible to
near-infrared). The pattern of change in both quantities mimics the
changes seen in reflectance values. The single-scattering albedo is
close to unity in the 0.8 to 4 micron domain with a large change at 2.8
microns due to the hydrated mineral absorption, and is much smaller in
both the shorter and longer wavelengths. The opposite pattern is seen
for the imaginary index. Thus, multispectral calculations are essential
in order to derive atmospheric heating rates and surface solar fluxes
with any accuracy.

References: [1] Mustard, J. F., and J. F. Bell III (1993) GRL, 21,
353-356. [2] Owen, T., and C. Sagan (1972) Icarus, 16, 557-568. [3]
Roush et al. (1992) Icarus, 99, 42-50. [4] Hapke, B., JGR 86, 3039-3054,
1981. [5] Hapke, B., Icarus, 67, 264-280, 1986. [6] Pollack, J.B., et
al. (1994) JGR, in press.

18.10-P Gillet V. Lellouch E.
Does Mars' Middle Atmosphere Really Have a Retrograde
Circulation Near Equinox?

Millimeter-wave observations of Mars in the J = 2-1 and J = 1-0
rotational lines of ^12CO (and in the J = 2-1 of ^l3CO) were performed,
at 115 and 230 Ghz respectively, in December 1990 and 199?, with the
IRAM 30m radiotelescope at Pico Veleta, Spain.

On December 20, 1990, the observations was performed shortly before
northern spring equinox (Ls = 352 degrees) at 25 Khz and 1 Mhz frequency
resolution. The december 29,1992 - January 2,1993 observations (four
consecutive nights) were obtained after northern equinox (Ls = 19
degrees) at 100 Khz and 1 Mhs frequency resolution. In both cases, the
Martian disk was modestly resolved (telescope beam diameter = 12-13
arcsec at 230 Ghz and Martian disk = 15 arcsec).

Direct absolute wind measurements in the middle atmosphere (near 50 km)
were obtained from Doppler shifts in the J = 2-1 CO transition at
230.538 Ghz, with an absolute uncertainty 1 sigma ~- 11-19 m/s for
December 90 observations and 1 sigma ~- 5-7 m/s for the December 92

Unexpectedly, from theoretical standpoint, the results lndicate, for
two observing periods and for all nights in 1992, independently, a
retrograde zonal wind. Preliminary modelling, assuming a solid rotator
like zonal circulation,.indicate a ~ 120 m/s equatorial velocity. In
relation with the dynamical picture, our data should also provide
information about possible hemispheric variations of Mars' thermal
structure. Finally, high signal-to-noise combined observations of ^l2CO
and ^13CO may also allow to retrieve the CO vertical profile at 0-70 km.

18.11-P Dason S. B. Conrath B. J. Pearl J. C.
A New Algorithm for Detecting Vertically Propagating Waves in
Planetary Atmospheres

Phenomena such as storms and flow over topography can produce vertically
propagating waves in planetary atmospheres. The ability to detect and
quantify these waves therefore contributes to our understanding of
atmospheric dynamics. We have developed algorithms to detect the
vertical temperature perturbations caused by these waves, using combined
spatial and spectral infrared limb data. We present an algorithm
designed to optimize the vertical resolution achievable from such data
sets. Outgoing radiance at the top of the atmosphere of Mars is modeled
using the radiative transfer equation. Optimization is achieved through
a particular choice of basis vectors for the kernel and a careful
selection of numerical grid. Direct application of this algorithm to
simulations of Mars Global Surveyor Thermal Emission Spectrometer data
shows that gravity waves with wavelengths as small as approximately 13
km. can be detected at pressures less than 1.25 mbars.

18.12-P McMillan W. W. Pearl J. C. Conrath B. J.
The Thermal Spectrum of Martian Dust from Mariner 9 IRIS
Emission Angle Pairs

We have investigated the dust suspended in the martian atmosphere for
portions of the Mariner 9 mission. Utilizing pairs of IRIS thermal
spectra of the same geographic location taken minutes apart but at
different emission angles, we have determined the spectral signature of
the dust from 300 to 600 cm^-1 and from 800 to 1200 cm^-1, as well as
the total dust opacity. Though limited in their temporal and spatial
coverage, there are sufficient numbers of pairs of spectra to
characterize the martian dust between orbits 68 and 176 (313 degrees <
L(sub)s < 343 degrees).

To first order, we find no appreciable change in the spectral structure
of suspended dust in the martian atmosphere between orbits 68 and 176.
The dust opacity at 9 microns decreased from tau ~ .22 on orbit 68 at 55
degrees S to tau ~ .16 on orbit 176 at 35 degrees S, and the column mass
of dust dropped from approximately 4 x 10^-5 gm/cm^2 to approximately 3
x 10^-5 gm/cm^2, respectively. Measured dust opacities are consistent
with the temporal and spatial distribution of dust we infer from the
heights of limb hazes determined from Mariner 9 TV observations
(Anderson and Leovy (1978) JAS, 35, 723); and with dust opacities for
different years but at the same martian seasons determined from Viking
IRTM observations (Martin (1993) JGR, 98, 10941). In addition, our
measured dust opacity for orbit 176 (L(sub)s ~ 343 degrees) is
comparable to that obtained by Santee and Crisp (JGR, 98, 3261, 1993)
from Mariner 9 IRIS spectra of the same latitude and similar L(sub)s.

This work was supported in part by the National Research Council and
NASA contract #NAS 5-31378.

18.13-P Petrosyan A. S. Karelsky K. V.
A Model for the Simulation of the Near Surface Phenomena on

A model for the study of the interaction of the planet surface with the
atmosphere is discussed. Wiscous atmosphere flows with any small
viscosity remain to be zero on solid boundary in consequence of the
no-slip conditions. Our prime interest here is with the dynamical
phenomena attendant on the wind blowing the Mars surface (dust
transportation, direct action of atmosphere on surface). With this in
mind we abandon the zero conditions for horizontal velocity in favor of
using non viscous equations. In this case corresponding velocity
component near the surface is forecasting value and total wind flow is
determined only by zero vertical component boundary conditions. We use
the system of univelocity equations of the multi-phase fluid for Mars
atmosphere description to achieve a better understanding of the
dependence of the structure and governing mechanisms of the mesoscale
processes of the thermal and orographic inhomogeneity and on dust
transport. The major advantage of using this particular type of
equations is in the fact, that allows to describe uniformly the
atmospheric flow along mountains, craters and sandy surfaces. This lets
us to analyze the structure of wind flows over complex terrain in the
important case of nonstationary condition of dust lifting and dust

This work was supported by Russian Space Agency.

18.14-P Vasilyev V. P.
Speed-Resonance in Planetary Atmospheres: Astronomical and
Environmental Aspects

The infrasound wave generation due to speed-resonance phenomenon for the
solar terminator moving through the Terrestrial and Martian atmospheres
are described. The power spectra of density and pressure wave
disturbances are obtained and analyzed with the point of view of their
experimental identification as local dynamic anomalies of atmospheric
parameters at sunrise and sunset. The unknown before environmental
significance of this phenomenon owing to bio-active infrasound wave
amplification in the industrial pollution regions at 30_50 deg
terrestrial latitude zones is discussed.

Wednesday, 1:30 - 2:15 Crystal Ballroom
R. L. Millis, Moderator

Wednesday, 2:15 - 3:00 Crystal Ballroom
B. L. Bjoraker, Moderator

18B.01-INV Encrenaz Th.*
Current Knowledge of Martian Atmospheric Composition

No abstract available.

SESSION 19 ....... Comets III
Wednesday, 3:30 - 5:30 Crystal Ballroom A
M. R. Combi and P. Eberhardt, Moderators

19.01 Levasseur-Regourd A. C.* Hadamcik E. Renard J. B. Worms J. C.
Evidence for Two Classes of Comets from Polarimetric Properties
at Large Phase Angles

Phase dependence of the linear polarization of scattered light is a clue
to physical properties of the dust. It allows comparisons between
various grains or regoliths, since no normalization with distance to the
Sun and to the Observer is needed. All the phase curves, derived from
appropriate fits within the domains where data are available, present
the same trends with an inversion angle near 20 degrees. From data
obtained for 21 comets, a dichotomy is pointed out at phase angles
greater than about 45 degrees (for any wavelength range in the visible
or near infrared domains) between comets with a polarization almost
equal to 30 percent near 90 degrees (e.g. West E- 1976 VI, Halley 1986
III, Bradfield 1987 XXIX, Levy 1990 XX), and other comets with a
polarization of about 15 percent near 90 degrees (e.g.
Kobayashi-Berger-Milon 1975 IX, Crommelin 1984 IV, Brorsen-Metcalf 1989
X, Austin 1990 V). This result is discussed in terms of cometary
activity, of temporal fluctuations in brightness, of gas to dust ratio,
and of orbital properties. The polarimetric properties at large phase
angles are tentatively interpreted through measurements performed on
dust samples (with various sizes, albedos, densities) during parabolic
flights on a plane dedicated to microgravity studies.

This work was partly supported by French Programme National de
Planetologie and by CNES.

19.02 Wehinger P. A.* Wyckoff S. Wagner A. Jury R. A.
Belton M. J. S. Tegler S.
Spectra of Comet P/Halley at Large Heliocentric Distances

Spectra of Comet P/Halley were acquired at a post-perihelion distance of
4.8 AU using the NOAO/CTIO 4-meter telescope, long slit spectrograph,
and photon-counting detector (2d-Frutti). Over the spectral range Lambda
Lambda = 3500-5500 Angstroms, detected emission features included: CN
(0,0) 3876 Angstroms, C3 4040 Angstroms, and C2 (0,0) 5165 Angstroms.
The spectra have been reduced in both 2-d and 1-d format, are flux
calibrated, and have a spectral resolution Delta Lambda ~ 6 Angstroms.

Additional spectra acquired at a post-perihelion distance of 8.0 AU show
no detectable emission features over the range Lambda Lambda = 3600-6000
Angstroms. These observations were obtained using the NOAO/KPNO 2.2-m
telescope, long slit spectrograph, and ICCD. At 8.0 AU the spectrum is
best characterized as scattered solar continuum. The angular extent of
the coma is ~30 arc sec, which corresponds to a linear diameter of 1.6 x
10^5 km at Delta = 7 AU.

The post-perihelion spectra at 4.8 AU are compared with pre-perihelion
spectra acquired at 4.8 and 4.7 AU with the 4.5-m Multiple Mirror
Telescope and an intensified Reticon (cf. Wyckoff et al. (1985) Nature,
316, 241, 1985). The post-perihelion spectra show CN production rates
that are ~15 times greater than in the pre-perihelion spectra at the
same heliocentric distance. This difference provides some measure of the
thermal inertia of the cometary nucleus.

This research is supported in part by the NASA Solar System Exploration
Division and by the NASA Space Grant College Program.

19.03 Hanner M. S.* Hackwell J. A. Russell R. W. Lynch D. K.
The Silicate Emission Feature in Comet Mueller 1993a and the
Origin of Annealed Grains

Since comets formed in cold regions of the solar nebula, unaltered
interstellar silicate grains should have survived in comets. Yet,
spectra of cometary dust differ from spectra of interstellar dust. A
distinct peak at 11.2 micrometers, attributed to crystalline olivine,
was detected in P/Halley and long period comets Bradfield 1987 XXIX and
Levy 1990 XX, but this signature of annealed grains is absent in spectra
of interstellar dust.

We have obtained 8 - 13 micrometers spectra of new comet Mueller 1993a
at R = 2 AU with the Aerospace Corp. broadband array spectrograph at the
NASA IRTF. The spectra clearly reveal the 11.2 micrometers peak for the
first time in a new comet. This result implies that cyrstalline olivine
must have been widespread in the solar nebula material from which the
comets formed, although it constitutes only a fraction of the silicate

Where did these annealed grains originate? The required temperature for
formation of crystalline grains in the solar nebula was attained at
small heliocentric distance, R < ~ 1 AU. This would imply that extensive
radial mixing occurred between the inner and outer solar nebula, in
contrast to some current nebula models. Either radial mixing was more
extensive than presently modeled or the olivine grains had a presolar
origin. If presolar, then the lack of an olivine signature in the
spectra of the interstellar medium and young stellar objects is

Supported at Jet Propulsion Laboratory, Cal Tech, under contract with
NASA and at Aerospace Corp. under the Sponsored Research Program.

19.04 DiSanti M. A.* Mumma M. J. Reuter D. C. Davies J. K.
Geballe T.
Systematic Observations of a New Cometary Organic Feature:
Heliocentric Development in Comet P/Swift-Tuttle

We present the first clear evidence for a new organic feature in comets,
which accounts for nearly one-half of the flux contained in the organic
emission near 3.4 ,micrometers.

Moderate resolution ( lambda/delta lambda ~ 500) infrared spectra of
comet P/Swift-Tuttle, covering ~3.2 - 3.6 micrometers, were obtained
using the Cryogenic Spectrometer (CRSP) on the Kitt Peak 4-m telescope,
as well as the Cryogenic Grating Spectrometer (CGS4) on the UKIRT
telescope on Mauna Kea. The observations were conducted UT 1992 November
9 and 12 (CRSP), and November 13 and 27 (CGS4). During this interval,
the heliocentric distance RHdecreased from 1.12 AU to 0.99 AU, while
delta increased from 1.16 AU to,l.31 AU. In addition, on UT November 26,
a CGS4 echelle spectrum (lambda/delta lambda ~ 8000), coveling 3.414-
3.441 ,micrometers, was obtained.

Upon removal of contributions from the dust continuum, and from
CH(sub)30H v(sub)2- and v(sub)9-band emissions (which comprise roughly
60% of the flux within the 3.4-micron cometary feature), the spectra
reveal a fairly symmetric?esidual emission, encompassing ~3.38 -3.47
,micrometers. This feature has a heliocentric variation indicative of
volatile species, a notion reinforced by the presence of a pronounced
peak at ~3.425 micrometers (2920 cm^-l), and broader "shoulders"
displaced ~20 cm^-1 to either side of the peak. Possible progenitor
candidates will be discussed.

19.05 Reber M.* Eberhardt P. Krankowsky D. Hodges R. R.
The Deuterium and 18-O Abundance in Water of Comet P/Halley

The high mass resolution of the double focussing M-Analyzer of the
Neutral Mass Spectrometer on the Giotto probe allowed the unambiguous
and accurate measurement of the abundance of the rare ions with masses
20 and 21 amu/e in the coma of comet P/Halley. These masses are
dominated by the molecular ions H(sub)2D^16O^+, H(sub)3^17O^+,
H(sub)2^18O^+, and H(sub)3^18O^+. From the preliminary evaluation of
these data estimates of the ^18O/^16O and D/H ratio in Halley's water
have been published (Eberhardt et al., A & A, ,187, 435, 1987). We have
carefully reevaluated these data, including extensive laboratory
calibrations of the instrument performance and detailed ion-chemical
modeling of the coma. We obtain an 18^O/^16O ratio with an uncertainty
of less than 10% which is compatible with the terrestrial ratio. The
deuterium in the water of comet P/Halley is distinctly enriched in
comparison with terrestrial water. The implications of this observed
high D/H ratio in P/Halley will be discussed.

19.06 Crovisier J.* Biver N. Bockelee-Morvan D. Colom P. Jorda L.
Lellouch E. Paubert G. Rocher P. Despois D.
P/Schwassmann-Wachmann 1: Radio Observations of Carbon Monoxide
and the Explanation of its Distant Activity

Observations of P/Schwassmann-Wachmann 1 at the IRAM 30-m radio
telescope confirm the CO J(2-1) line previously reported by Senay and
Jewitt (IAU Circ. No 5929) from observations at the JCMT.

The line shape, obtained with a velocity resolution of 0.06 km s^-1,
shows a very narrow component of 0.6 K brightness temperature and 0.15
km s^-1 wide, corresponding to a cold kinetic temperature (about 10 K).
A broader component of 0.1 K, extending from -0.9 to 0.5 km s^-1, is
also present. A contamination of this broader component by galactic CO
emission gradients cannot be excluded at the present time.

The narrow line is found to be blueshifted by 0.5 km s^-1. This is an
evidence of anisotropic outgassing from the sunlit side of the nucleus.
The blueshift is stronger than the expansion velocity expected from the
sublimation of exposed pure CO ice. It suggests that CO is percolating
through a warmer dust mantle or porous ice.

The intensity of the narrow line corresponds to a CO production rate of
1 x 10^28 molecules s^-1 or more, depending upon the true rotational
distribution of the molecule. This is sufficient to explain why this
comet shows persistent activity at 6 AU from the Sun.

19.07 Senay M. C.* Jewitt D.
Temporal Variability of CO Production from Comet
P/Schwassmann-Wachmann 1

Comet P/Schwassmann-Wachmann 1 is famous for its episodic outbursts of
activity, characterized in the optical by sudden continuum brightening
and the development of CO+ emission lines. The agent responsible for
this activity has until recently eluded detection. We have identified CO
as the primary activity driver in comet P/Schwassmann-Wachmann 1. Our
observations of CO J(2-1) and J(3-2) submillimeter emission (see Fig),
obtained with the JCMT, show that the CO production rate is
approximately 2000 kgs^1 in quiescence, sufficient to account for the
previously observed features of optical activity. The Doppler shift of
the J(2-1) and J(3-2) lines indicates the CO sublimates from near the
subsolar point on the nucleus and forms a jet of gas and dust that moves
outwards from the comet at a speed consistent with this mode of
creation. The CO production rate as measured by us is also vanable, as
it should be if CO is the driver of the continually changing levels of
optical activity. We find that the CO production rate can change by
factors of two or more on relatively short time scales.

19.08 Meier R.* A'Hearn M. F.
Monte Carlo Simulations of the Sulfur Triplet at 1814 Angstrom
in Optical Deep Cometary Comae--A Comparison with UV

Measured column densities and line ratios of the sulfur ^3S^O --> ^3P
transitions indicate that even in a moderately dense cometary coma the
SI triplet at 1814 Angstroms is impaired by opacity effects (cf. Azoulay
and Festou, in ACM II, Uppsala, 273, 1986). Recent observations of
comets recorded by the Faint Object Spectrograph (FOS) aboard the Hubble
Space Telescope (HST) show similar depth effects for the sulfur lines as
do earlier data of the International Ultraviolet Explorer (IUE).

A Monte Carlo model is presented that determines line intensities of the
^3S^O --> ^3P sulfur transitions in an optical deep coma. In our
radiation transfer model we consider multiple scattering including the
Greenstein effect. The velocity space of the gas is described by an
isotropic 3-dimensional and Maxwellian-distributed random component and
a radial, constant bulk component. A frequency redistribution function
accounts for changes in frequency and angle between absorbed and
reemitted photons. Our velocity space assumes locally thermalized sulfur
at equal radial and transverse temperature. This is an appropriate
approximation of the inner coma where both attenuation of solar or
scattered photons and multiple scattering become important. Spatial
distributions of sulfur we derive from Haser models using effective
scale lengths. Essentially we are able to model averaged (apparent)
column densities for any given shape of aperture, and for any radial
density and radial temperature profile. Simple hemispherical comae are
investigated to study the effect of spatial asymmetries. Model results
are compared with FOS and IUE observations of S, CS, and OH.

This work was supported by the Swiss National Science Foundation
(fellowship for R.M.) and NASA grant NAGW 902.

19.09 Cruikshank D. P.* Roush T. L. Bartholomew M. J.
Organic Solids on the Surface of 5145 Pholus

We have continued the study of the near-IR spectrum of 5145 Pholus begun
with the acquisition of the improved data set discussed by Cruikshank et
al. (B.A.A.S. 25, 1125, 1993). We have computed Hapke scattering models
of the surface of Pholus using optical constants derived by us from
reflectance spectra of high-bitumin organic solids of terrestrial
origin, and optical constants of other materials derived by various
investigators. While details in the spectrum of Pholus remain to be
matched by realistic models, we report on the progress we have made in
applying multi-component scattering models using plausible materials
expected on objects recently arrived in the planetary region from the
Kuiper Belt, as Pholus is thought to be.

Research supported by NASA Planetary Astronomy program.

19.10 Moore M. H.* Ferrante R. F. Nuth J. A. III
Infrared Study of Laboratory Ices Containing Methanol

Water ice is a major constituent of comets. A variety of other simple
molecules have also been identified in the condensed phase (e.g. C0,
CO2, CH30H, H2CO, etc), along with an appreciable amount of silicates.
To understand some of the physico-chemical and radiation-chemical
processes which might affect these ices and ice/silicate composites,
appropriate laboratory experiments are required.

Current investigations include the study of ices such as CH30H and H20 +
CH30H. Laboratory infrared spectroscopy is a used as a tool to identify
new species synthesized in the ice as the result of proton irradiation.,
mass analysis of volatile species released from the processed ice during
slow warming is combined with information from the infrared spectra of
the remaining ice. Infrared spectra of any non-volatile residue are
measured. Comparison of results for these ices deposited as thin films
on aluminum and on porous silicate smokes are in progress. Silicate
smokes are used as simple grain analogues. Our laboratory smoke deposits
form as amorphous structures w1th a submicron grain size. Previous
experiments showed that mostly crystalline (phase) ices form on these
smokes at T < 20 K, a result different from the amorphous phase formed
on aluminum substrates. Effects of silicates on the spectra or
vaporization of the ices or on the nature of the residue will be
presented. Application of results to comets will be discussed.

* This work was supported by the NASA Planetary Atmospheres Program
**Supported by the NASA Planetary Materials and Geochemistry Pro?ram

19.11 Samarasinha N. H.* Klavetter J. J.
Simulations of CN Jets in Comet Halley

We have numerically simulated the CN jets of comet Halley to model the
radial velocity of jets. For consistency, the techniques used by
Klavetter and A'Hearn (1992) to analyze the observed CN jets were
adopted in analyzing the simulated images.

The preliminary results indicate an enhancement of the radial jet
velocities compared to that of the jet material. Thus, the jet velocity
of 1.7 km/sec obtained by Klavetter and A'Hearn is most likely an
enhanced value for the outflow velocity of the jet material. We will
present the comparisons between the observations and simulations and
constrain the associated physical properties of the jet materials and
the jets themselves.

19.12 Moreels G.* Clairemidi J. Goidet-Devel B. Rousselot P.
Detection of Cometary Polycyclic Hydrocarbons: A Possible
Target for Rosetta

Polycyclic aromatic hydrocarbons are now widely considered as an
important component of the interstellar medium. Some of them have been
identified in the solar system: in meteorites (Murchison chondrite) and
in tholins obtained in simulated atmospheres of Titan and Jupiter.
Evidence of the detection of phenanthrene C(sub)14H(sub)10 in Halley's
comet is presented here. Observational data consist of the spectra taken
from the Vega 2 spaceprobe in the near-UV range. The emergence of a band
at 342-375 nm with four peaks at 347, 356, 364 and 374 nm is assigned to
a fluorescence band of phenanthrene in the S(sub)-->1S(sub)0 electronic
transition. The identification is based on a comparison with a
laboratory spectrum obtained in a laboratory laser induced fluorescence
experiment. Another band at 323 nm is tentatively identified as due to
naphthalene C(sub)10H(sub)8 on the basis, also, of a comparison with a
laboratory spectrum of naphthalene under jet-cooled conditions.

The presence of aromatic hydrocarbons in comet Halley provides an
additional argument showing that primitive bodies of the solar system
contain organic material. It emphasizes the scientific interest of the
Rosetta mission.

Spectro-imagery in the UV and near-UV parts of the spectrum is a most
powerful method for detecting aromatic hydrocarbons and other organic
compounds that show fluorescence bands in this region.

SESSION 20 ....... Io
Wednesday, 3:30 - 5:30 Crystal Ballroom B
D. L. Blaney and N. Thomas, Moderators

20.01 Davies A. G.*
Silicate Eruption Dynamics on Io: Comparison with Observations

Infrared outbursts on Io are of volcanic origin. The implied
temperatures are indicative of silicate, rather than sulphur volcanism.
Changes in temperature, area and power output from the outburst of
Janualy 9th 1990 provide other constraints on the range of eruptive
processes taking place on Io. Modelling of this event produces an
eruption rate of similar size to terrestrial flood basalt eruption
episodes. Using this mass eruption rate as a starting point, the ascent
and eruption of silicate magma is modelled, with sulphur as one of a
number of possible volatile components. The volcanological model yields
emption parameters such as degree of fragmentation of the magma,
pressure in the vent and eruption velocity. The conduit radius (or
fissure half-width) is also determined, although this may change through
erosion of the vent. This process also reduces eruption velocity as the
flow expands laterally. Possible ballistic ranges of associated
pyroclastic ejecta are calculated. The volatile content of the silicate
material strongly effects eruption behavior, especially eruption

Modelling of flow formation and cooling of the erupted material is in

This work was partially supported by a grant from the Science and
Engineering Research Council, UK, and partially performed while the
author held a National Research Council-JPL Research Associateship.

20.02 Howell R. R.*
The Nature of the Volcanic Activity at Loki

The nature of the volcanic activity on Io remains uncertain, despite P
detailed observations from the Voyager spacecraft and extensive
monitoring from earth. However, the monitoring program has provided two
types of new information concerning the volcanic activity at Loki.
First, the combined data set from the many ground-based observers
suggests that Loki follows a cycle where it remains relatively inactive
for one or two years, then brightens considerably for a period of a few
months. Second, the mutual event observations during 1991 provide the
highest spatial resolution data, which clearly resolve the emitting

The availability of these new types of data has made it worthwhile to
reinvestigate the models for thermal emission from silicate flows. New
simple numerical models of the expected emission provide a very good
match to the Voyager IRIS spectrum, and also seem consistent with the
constrains imposed by the ground-based observations. With some
reasonable simplifying assumptions it is also possible to develop
analytical expressions for the expected flux and spectral shape. These
analytical E models provide additional insight into which physical
parameters of the volcanic activity can in fact be constrained by the
infrared observations.

This work was supported by the NASA Planetary Astronomy Program under
grant NAGW-1276.

20.03 Kargel J. S.*
Major Roles of P and Other Minor Elements in Io's Sulfurous

The observed constituents of the Io torus are O>S>> Na>K, of which the
first three are probably abundant on Io (A.F. Cheng, 1984, J. Geophys.
Res., v. 89, p. 3939-3944; N.M. Schneider et al., 1991, Science, v. 253,
p. 1394-1397). Elemental S, S0(sub)2, Na(sub)2S(sub)5, Na(sub)20,
Na(sub)2S0(sub)4, and H(sub)2S may be the major molecules on Io (D.C.
Pieri et al., 1984, Icarus, v. 60, p. 685-700; D.B. Chrisey et al.,
1988, Icarus, v. 75, p. 233-244; R.R. Howell, 1989, Icarus, v. 78, p.
27-37; D.B. Nash and R.R. Howell, 1989, Science, v. 244, p. 454-457;
D.B. Nash and G.R. VanHecke, 1992, Lun. Planet. Sci. Conf., p. 965-966).
S0(sub)2 might constitute a widespread but very thin volcanic sublimate
over a much thicker crustal layer dominated by O-poor sulfurous lavas,
including elemental S. Cosmic abundances and basic elemental properties
suggest that Ionian lavas and other surface materials may include
important quantities of P, Cl, K, and H and traces of Rb, Cs, As, Se,
Te, F, Br, and I, all of which either behave geochemically like S or
form strong covalent bonds with it. Despite the multiplicity of complex
geochemical processes operating on Earth and other planets, these
elements all tend to be partitioned with S. This is reflected by the
fact that the elements listed above commonly are concentrated together
in highly diverse geologic materials. One expects that similar
concentrations of these elements would occur on Io's surface. The cosmic
(CI chondrite) abundances of some of these elements are significant
relative to S and Na (Na/S = 0.093, P/S = 0.022, Cl/S = 0.013, and K/S =
0.011; E. Anders and N. Grevesse, 1989, Geochim. Cosmochim. Acta, v. 53,
p. 197-214). Thus, it is desirable to consider possible effects of minor
constituents on the properties of sulfurous magmas. Although there are
major gaps in our knowledge of the pertinent chemical systems, insights
can be obtained from the system S-Na-K-P (this consideration omits O as
an important constituent). Sodium and potassium form sulfides (e.g.,
Na(sub)2S(sub)5, K(sub)2S(sub)6) that are virtually immiscible in molten
S; thus, phase equilibria can be considered in the separate systems
Na(sub)2S(sub)5-K(sub)2S(sub)6 and S-P, which may constitute two
distinct series of Ionian lavas. The system S-P has a eutectic near 7.5%
P, which crystallizes to 27% P(sub)4S(sub)10 + 73% S(sub)8. The eutectic
melting point (353 K) is substantially lower than sulfur's (391 K). If
elemental S is present in excess over the eutectic ratio of S:P then
pure S magmas may also form; but the lower melting point of the S-P
eutectic should make it more igneously active than pure S. Besides
affecting the melting~point of S, small amounts of P markedly influence
the spectroscopic characteristics, elasticity, and visosity of solid and
molten S. Alkali sulfides melt at higher temperatures than S and S-P
(e.g., the peritectic m.p. of Na(sub)2S(sub)5 ~ 531 K). Oxygen, a major
element on Io, and small quantities of Cl and other elements no doubt
would further complicate the physical chemistry. Indeed, Kargel (1993,
Lunar Planet. Sci. Conf. XXIV, p. 751-752) suggested that Io has a thick
lower crust of carbonates and sulfates (m.p. 750-1200 K) beneath a few
kilometers of S. Active extrusions of the predicted magmas, including
S-P, Na(sub)2S(sub)5-K(sub)2S(sub)6, and carbonate-sulfate, may best
explain Io's thermal emission spectrum.

20.04 Buratti B. J.* Mosher J. A. Terrile R. J.
Does SO2 Frost Condense on the Nightside of Io?

The surface of Io is believed to consist primarily of elemental sulfur
allotropes and S02. The exact nature of these materials, their purity,
their fractional surface coverage, and their interaction with a possible
S02 atmosphere are still uncertain.

During the Voyager 2 encounter several sequences of the dark
(Jupiter-illuminated) side of Io were obtained. Because it is not
possible to obtain such observations from Earth these images represent a
unique opportunity for understanding several problems relating to the
surface composition of Io. One such problem is whether S02 frost
condenses on the nightside of Io. According to standard vapor pressure
curves (Matson and Nash, 1983), the vapor pressure of S02 drops more
than 10 orders of magnitude between day and night temperatures on Io,
which differ by 70K. The existence or absence of S02 condensates on the
night side is thus a sensitive indicator of an S02 atmosphere. A recent
analysis by Simonelli et al. (1994) failed to show any local
brightenings on the dark side in the Voyager violet filter, which is
sensitive to the presence of S02 frost.

Our analysis applies a new approach to this problem. Instead of
searching for albedo changes, we search for color changes on the dark
side of Io. Because S02 frost is bluer than elemental sulfur in the UV
-violet portion of the spectrum, the nightside of Io should be bluer
than the dayside if S02 frost is condensing on it. A color mosaic of the
dark side of Io, created from ultraviolet and green images from the
Voyager 2 spacecraft, shows that in fact the darkside of Io is bluer
than the same region und Rr solar illumination. This result provides
evidence for the global condensation of S02 frost on the dark hemisphere
of Io. Performed under contract to NASA.

Matson, D. and Nash, D. (1983) J.G.R. 88, 4771. Simonelli, D. et al.
(1994) Icarus 107, 375.

20.05 Spencer J. R.* McGrath M. A. Sartoretti P. McEwen A. S.
Post-Repair HST Imaging of Io: Surface Changes and Spectral

In March 1994 we imaged all longitudes of Io in sunlight at 0.34, 0.41,
0.56, 0.78, and 1.02 micrometers with WFPC2 on the repaired HST, with
high S/N and 22 pixels across Io's disk, producing the best map of Io's
albedo patterns since Voyager. At Voyager wavelengths, changes since
1979 are numerous and widespread but generally subtle at 200 km
resolution. Changes have occurred near the volcanos Loki, Masubi, and
Kanehekili, but other changes, most notably a darkening of southern
Colchis Regio also seen in 1993 images by Sartoretti et al. (1994), are
not correlated with known volcanic centers. This dataset includes the
first images of Io's albedo patterns in the near-IR, not seen by
Voyager. The 0.78 / 0.56 micrometers reflectance ratio, which maps the
strength of the 0.6 - 0.7 micrometers absorption edge seen in Io's
disk-integrated spectrum, is unusually high in a region coinciding
exactly with the pyroclastic deposits of the Pele volcano. The material
responsible for this absorption edge, possibly S(sub)20 or Na(sub)2S
(Hapke 1979, Nash 1993), is thus being generated in unusually large
concentrations by Pele. The 0.78 / 0.56 micrometers ratio is also high
in the polar regions. At 200 km resolution the 0.78 and 1.-02
micrometers appearance of Io is essentially identical. We also obtained
FOC images of Io at 0.34 micrometers, with somewhat higher resolution
but poorer S/N. None of the images show gbvious eruptive plumes on the
limb, though analysis is continuing.

We also obtained a 13-minute exposure 1.04 micrometers image in Jupiter
eclipse on March 6 1994, to look for high-temperature volcanic thermal
emission. None was seen, providing an upper limit to the surface area of
exposed silicate magma of about 0.1 km^2 at any Io hot spot on that

20.06 Ballester G. E.* Clarke J. Trauger J. Stapelfeldt K.
Crisp D. WFPC2 Investigation Definition Team Strobel D.
McGrath M. Schneider N. Ajello J. Combi M.
Ultraviolet Observations of Io with HST: WFPC2 Imaging and GHRS
and FOS Spectroscopy

A number of ultraviolet observations of Io with tne Hubble Space
Telescope were made in June 1994 to study the main constituents of Io's
atmosphere, molecular SO2 and neutral oxygen and sulfur, and the surface
reflectivity. Images were obtained with the Wide Field Planetary Camera
2 (WFPC2) as part of the GTO program. A separate guest observer (GO)
program was dedicated to disk-integrated spectroscopic observations with
the Faint Object Spectrograph (FOS), and spectroscopic mapping with the
Goddard High Resolution Spectrograph (GHRS). Images with a series of
filters centered at ~2180, 2550, 3360, and 3800 Angstroms show the
distribution of the surface reflectance and also of atmospheric SO2 gas
absorption. Disk-integrated spectra were obtained with the FOS that
measure the SO2 atmospheric abundance from the ~2100 Angstrom absorption
signature. Comparison of these two sets of data should add to their
value in determining the spatial distribution of this gas. Io's neutral
O and S components were studied with spatial scans of their far-UV
emissions with the GHRS. These scans should provide valuable information
on the spatial distribution of the emissions. East/west and north/south
scans of both optically thick and optically thin emissions were made,
and depending on the details of the observing sequences, complimentary
information may be obtained on the atmosphere and the dependence of the
emissions of the Io plasma torus. WFPC2 images of Io were also obtained
using the sodium Wood's filter covering ~1150-2100 Angstroms (with very
low red-leak) for detection of neutral O and S far-UV emission features.
The far-UV sunlight reflected by the surface is also measured in these
images. Preliminary results will be presented. This work was supported
by STScI grant GO 5438.01-93A and JPL grant 959122.

20.07 Lellouch E.* Strobel D. Belton M. Paubert G. Ballester G.
de Pater I.
Millimeter Wave Observations of Io's Atmosphere: New Data and
New Models

The analysis of several rotational lines of SO(sub)2 in Io's atmosphere
has indicated a relatively dense (3-40 nanobar), patchy (5-20 % of the
projected surface) and hot (600 K near 1 nb) atmosphere, presumably of
volcanic origin (Lellouch et al. 1992). However, elaborate
radiative-conductive models (Strobel et al. 1994) fail to explain such a
high temperature and HST data (Ballester et al. 1994) yield atmospheric
pressu?es in the range (0.1-3 nb). The latter two authors have proposed
that perhaps the width of the millimeter-wave lines, rather than being
purely thermal, includes a velocity dispersion associated with the
dynamics of volcanic plumes.

This paper quantitatively explores this possibility. A Monte Carlo
ballistic simulation (Strom and Schneider 1982) of a plume is developed,
in which the free parameters are the ejection velocity and the cone
angle. This simulation generates the density and velocity distribution
of the plume gas, and the radiative transfer equation is then solved for
SO(sub)2 line emission from the plume gas.

The results will be presented and compared to new, higher S/N millimeter
wave observations of Io's atmosphere obtained in 1993 and 1994.
Highlights of these observations are (i) SO(sub)2 strong lines are
always present when observed and remain very similar on a long term
basis; (ii) intrinsically weaker lines can be detected on the leading
side than on the trailing side; (iii) the strong lines generally show a
red shift from their rest frequency; and (iv) the tentative detection of
SO at 219.949 GHz. These results will be interpreted in the context of
both the "old" and "new" models.

20.08 Trauger J. T.* Stapelfeldt K. R. Evans R. W. Crisp D.
Clarke J. T. Ballester G. E. WFPC2 Science Team
An HST View of Io's Corona

Io's corona has been observed in NaI and [SII] emission lines with the
WFPC2 and HST during a period of about three hours on 20 April 1994,
during which Io moved from 256 degrees to 290 degrees in orbital phase
and from 25 degrees to 125 degrees in System III longitude. Direct
images of Io were obtained in the F588N and F673N filters, then
spatially registered and subtracted to show the emission morphology from
~0.2 to several Io radii from Io's surface. Structure is seen to change
on spatial scales comparable to Io's diameter as Io moves from the
plasma confinement equator to 6 degrees north of the equator, and as Io
moves in orbital phase near Western elongation. These images complement
an extensive program of ground-based observations of Io-torus
interactions, extending our insight to small spatial scales near Io
where ion-neutral and charge-exchange collisions are concentrated.

20.09 Marconi M. L.* Smyth W. H.
Hybrid Fluid/Kinetic-Theory Approach to Modeling Io's

The atmosphere of Io, and any other planet for that matter, can be
divided into three physical regimes or regions: Region I closest to the
surface with Knudsen number <0.1, Region II with Knudsen number in the
range >0.1 and <10, and finally Region III with Knudsen number >10. The
significance of such a categorization is that in Region I the atmosphere
is of sufficient density (i.e., neutral-neutral collisions are
suff1ciently frequent) to permit an accurate description in terms of the
usual fluid formalism. In Region II, however, where for Io the impacting
torus plasma deposits most of its energy and momentum, the
neutral-neutral collision rate is large enough to strongly affect the
neutral velocity distribution function, but not large enough to enforce
local equilibrium, so that the fluid approximation is no longer valid.
As a result one must appeal to the full collisional nonlinear Boltzmann
equation (i.e., a kinetic theory approach) for a proper description of
Region II and its interaction with the torus. Finally in Region III
neutral-neutral collisions have become so rare that they may be ignored.
In this regime the atmosphere is determined by the atmosphere in Region
II together with collisions with the torus particles, and hence
the-linear Boltzmann description is appropriate. We are developing
hybrid fluid-kinetic models on massively parallel computers in order to
properly describe planetary atmospheres and their interactions with
impingmg sources of energy and momentum starting with Io. Here
preliminary results for Io's atmosphere are presented. A spherically
symmetric fluid model for Region I is combined with a spherically
symmetric DSMC (Direct Simulation Monte Carlo) kinetic theory model for
Region II. The atmosphere in Region III is also calculated according to
the scheme of Smyth and Combi (1988, Ap. J., 66, 888-918). The initial
model results presented will treat the effect of the torus interaction
in a simple fashion and will ignore chemistry. The overall structure of
the atmosphere and how it differs from an all fluid model will be
discussed as will future directions of development.

20.10 Wilson J. K.* Schneider N. M.
Variability of Io's Fast Sodium

Time variability in Io's atmosphere primarily reflects the processes
which produce the atmosphere. Surface-sputtered or frost-sublimation
atmospheres are expected to vary in density with a regular frequency,
due to either changing local plasma density or changing solar heating on
the trailing hemisphere. A volcanic atmosphere should vary less
regularly, corresponding to the activity of volcanic vents.

Io's neutral clouds may provide a means of measuring the atmospheric
variability. The neutral clouds are supplied by Io's atmosphere, and can
be studied with ground-based observations. The 'slow' sodium b-cloud,
the most consistently observed component of the sodium cloud, has
remained fairly stable over time^1. Other 'fast' components of the
sodium cloud have been shown to vary significantly in emission
brightness and morphology from year-to-year^2,3. These fast sodium cloud
features may provide a means of studying both long-term and short-term
temporal changes in Io's atmosphere.

We are undertaking a detailed analysis of over 200 near-Jupiter sodiun?
emission images taken from 1990 to 1992.We will present our measurements
of the sodium cloud, and discuss the implications for changes in the
atmospheric density, composition, and interaction with the local plasma.

This work has been supported by NASA grants NAGW-2484, NAGW-2492, and

[1] N. Schneider, W. Smyth, M. McGrath. 1989. in Time- Variable
Phenomena in the Jovian System, p. 75. [2] B. Flynn, M. Mendillo, J.
Baumgardner, 1994. JGR, 99, p. 8403. [3] J. ?;?Tilson, N. Schneider,
1994. Accepted by Icarus.

20.11 Brown M. E.*
The Structure and Variability of the Io Plasma Torus

Using G months of nearly continuous high-resolution spectroscopic
observations of [SII] 6716, 6731 Angstrom emission from the Io plasma
torus, I construct models of the average behavior of the torus and
explore periodic and non-periodic deviations from this average. The
observations were taken from December 1991 to June 1992 using the 0.6-m
Lick Observatory Coude Auxiliary Telescope coupled to the Hamilton
Echelle Spectrograph. Each spectrum is taken with the ~18 Jovian radii
spectral slit centered on Jupiter and aligned along the torus
centrifugal equator. Because of this very long slit, the spectra
simultaneously cover the dawn and dusk sides of the torus, and the
spectrum of Jupiter on the same slit allows precise calibration of
distances, intensities, and rotation velocities in the torus.

Parameters of the torus model include S^+ density, electron density, ion
temperature, and rotation velocity, all as a function of distance from
Jupiter and for both the dawn and dusk sides of Jupiter. I will compare
the model to the recent Voyager model of the torus and previous
ground-bascd models and discuss the implications of the observed radial
structure of the torus

Variations from the average behavior of the torus have the potential to
reveal much about the important physical processes governing the torus.
I will discuss persistent System III variations in densities,
temperatures, and velocities, as well as non-System III periodicites and
non-periodic behavior.

Similar observations of the torus are planned for approximately a month
surrounding the impact of Comet Shoemaker-Levy 9 with Jupiter this July.
Available results will be discussed.

20.12 Taylor M. H.* Schneider N. M. Bagenal F.
Peeling the Io Torus

Computer simulation of extensive ground-based torus images allow the
most detailed three-dimensional model of the Io torus to date. Previous
investigations have been hampered by low spatial resolution and
line-of-sight averaging through regions of the torus where parameters
change rapidly. A computer program called CITEP (for Colorado Io Torus
Emissions Package) was created to systematically remove the
line-of-sight averaging inherent in the images. CITEP simulates torus
images using variable plasma parameters, up-to-date atomic data, and
accurate viewing geometry.

We simulate the observations of Schneider and Trauger (submitted to Ap.
J., 1994) using the Voyager empirical torus model (Bagenal, J. Geophys.
Res., 99, 11043, 1994) as a base. We adjust the plasma parameters to
match torus emissions in layers from the outside moving inwards, each
time subtracting the modelled emission from the data. All the remaining
emission then originates from interior layers, which can be modelled and
removed in successive iterations. In this manner we can peel away
successive layers of the torus, revealing the radial and latitudinal
structure. By analyzing several images taken at different longitudes we
obtain the full, three-dimensional structure.

The 3-d structure provides critical insights into the torus mass and
energy balance and transport. For example, the results may help
determine the cause or causes of observed longitudinal variations. The
images do not uniquely constrain all plasma parameters, so the
limitations of our solution will also be discussed. This work was
supported by NASA grants NAGW-2484, NAGW-2492, and NASA/JPL contract

20.13-P Betts B. H. Nash D. B.
New SO2 Band Identifications in Io IR Spectra Based on
Laboratory Data

Comparing our lab spectra [1] with published telescopic spectra of Io
allows us to make new or alternative SO2 band identifications in the Io
spectra. As lab SO2 spectra have improved and expanded in both
reflectance (e.g., [1]) and transmission (e.g., [2]), it is becoming
apparent that SO2 by itself can explain most features in Io IR spectra.
Even bands that have been assigned to other materials, such as H20
(e.g., [3]), CO2 [4,5], and H2S [6], occur at the same wavelengths as
SO2 features, as discussed by [7], [2], and [1], respectively. Thus,
arguments for non-S02 materials on Io must rely on comparisons of shapes
and depths of bands, rather than just wavelengths. Here, we discuss
bands that were either not identified, not assigned, or likely
incorrectly assigned in their original publication. [8] identify a 2.55
+/- 0.1 micrometer band in their data, and [9] tentatively identify the
blueward part of this band in their data, but neither group assigned it
to SO2. We identify this band as the 3 nu(sub)3 (2.54 micrometers) band
of solid SO2; its redward portion seems to appear strongly in the data
of [3], although it was not explicitly reported. Although not
identified, the 2 nu(sub)1 + nu(sub)3 (2.79 micrometers) band of solid
SO2 appears in the data of [8], and is observed as a strong band in the
higher spectral resolution data of [3] . The band they observed is due
primarily and perhaps entirely to 2 nu(sub)1 + nu(sub)3 of solid SO2
[7], although there may be a contribution from a very small fraction of
water ice [3]. We identify and assign for the first time the
2.92-micrometer (3 nu(sub)1) solid SO2 band in the data of [3] and [10].
We tentatively identify and assign the 3.36-micrometer (nu(sub)1 +
nu(sub)2+ nu(sub)3) band in the data of [3] and [10]. Also apparent,
although weak, in the data of [10] are the 3.56-micrometer (2 nu(sub)1+
nu(sub)2) and 3.67 (a portion of 2 nu(sub)3) bands, as well as the major
4.07-micrometer (nu(sub)1 + nu(sub)3) and 3.78-micrometer (2 nu(sub)3)
bands. [11] identified two unknown bands in their Io spectra at 3.85 and
3.91 micrometers that were subsequently atributed to H2S by [6]. Our lab
data for SO2 slab ice has fairly weak bands at 3.85 and 3.92
micrometers, which were not clearly observed in our frost spectra. We
identify [1] these bands as the nu(sub)1 + nu(sub)3 + R(sub)x (R(sub)x
is a lattice rotational component) and nu(sub)1 + nu(sub)3 + T(sub)x
(T(sub)x is a lattice translational component) bands based on [2], who
identified similar bands in their lab data. On Io, any thick, solid SO2
should produce these bands, and thus SO2 may contribute to the bands
observed by [11], although H2S may contribute as well.

References: [1] Nash D. B. and Betts B. H. (1994) Icarus, in review. [2]
Schmitt B. et al. (1994) Icarus, in review. [3] Salama F. et al. (1994)
Icarus 107, 413-417. [4] Trafton L. M. et al. (1991) Icarus, 89,
264-276. [5] Sandford S. A. et al. (1991) Icarus, 91, 125-144. [6] Nash
D. and Howell R. (1989) Science, 244, 454-457. [7] Nash D. (1994)
Icarus, 107, 418-421. [8] Pollack J. et al. (1978) Icarus, 36, 271-303.
[9] Clark R. and McCord T. (1980) Icarus, 41, 323-339. [10] Cruikshank
D. et al. (1978) Astrophys. J., 225, L89-L92. [I 1] Howell R. et al.
(1989) Icarus, 78, 27-37.

20.14-P Nash D. B. Betts B. H.
Simulations of NIMS SO2 Spectra: Implications for Mapping SO2
on Io

We have measured IR reflectance spectra (2.3-23 micrometers) in the lab
for various phases of SO2 including frosts of various thicknesses [1].
The infrared spectrometer (NIMS) on the Galileo spacecraft will measure
spectra of Io in the range 0.7-5.2 micrometers. We have created
simulated NIMS SO2 spectra by convolving the approximate NIMS spectral
bandpass for each channel with our lab spectra for each of the standard
NIMS modes. Our simulations indicate NIMS will be able to distinguish
and map the distribution of various S02 phases on Io's surface (and also
identify such phases on other Galilean satellites, if detectable
quantities exist there), and assess the thickness of SO2 deposits. Our
results indicate NIMS will easily resolve major bands in SO2 frost or
ice, particularly when using all 408 channels (see figure, which appears
in the hard copy), but not isotope bands near 4.1 micrometers. The
apparent band positions, depths, and even the distinct presence of bands
vary considerably depending on the number of NIMS channels used. The
other critical factor for determining shape of the NIMS simulated
spectra, parlicularly for low numbers of channels, is the choice of a
point to "anchor" the spectrum. This is equivalent to the actual NIMS
choice of the starting grating position for each spectrum. This
drastically affects the shape, depth, and even detectability of bands.
Bands at 2.54, 2.79, and 3.78 micrometers in reflectance spectra are not
observed in thin frosts in the lab, presumably due to insufficient path
length; however, they do appear in thick frosts (~2 mm). These bands
appear in telescopic spectra of Io, indicating thick deposits do exist
on Io. However, groundbased data unavoidably favor the equatorial
regions, and spatial distribution of SO2 deposits and their thicknesses
can only be constrained slightly. Our simulated NIMS spectra indicate
that NIMS in its highest spectral resolution modes will be capable of
easily detecting the presence or absence and strengths of these bands,
and thus can map the thickness of SO2 deposits on
Io at much higher spatial resolutions than
currently possible. Thicknesses using these bands can also be mapped
with the low-resolution NIMS 17 and 51-channel
modes, but only by careful and consistent choices of anchor point to
optimize the data for the most significant bands.

References: [1] Nash D. B. and Betts B. H. (1994) Icarus, in review.

20.15-P Zhu X. Strobel D. F. Summers M. E.
On the Vertical Thermal Structure of Io's Atmosphere

A radiative-thermal conduction model for the vertical thermal structure
of Io's atmosphere is developed with solar heating by S02 absorption in
UV and near-IR bands and non-LTE cooling by S02 vibrational bands and
rotational lines. The model predicts the existence of a mesopause in
Io's atmosphere when the surface pressure exceeds about 10 nbar. The
radiative time constant for establishing a mesosphere/mesopause on Io is
only about 20 min, whereas the thermospheric radiative time constant is
about 1 hr. The asymptotic thermospheric temperature is about 270 K only
140 K greater than the surface temperature because at high altitudes
non-LTE cooling by S02 rotation lines exceeds cooling in the v2
vibrational band. Solar heating only models are incapable of generating
warm enough atmospheres to satisfy the observational inferences from UV
and especially millimeter-wave measurements. Joule heating driven by the
penetration of Jupiter's corotational electric field into Io's
conducting ionosphere is demonstrated to be the dominant heating
mechanism in the sub-nanobar regions of Io's atmosphere with
temperatures ranging from 150 to 1000 K as a function of decreasing
pressure from 1 to 0.1 nbar. The asymptotic thermospheric temperature
can attain a value as high as 1800 K.

20.16-P Sauter L. M. Schneider N. M. Spencer J. R.
Search for [SI] Emission from Io's Atmosphere

We report observations searching for [SI] 10825 angstrom emission from
Io's atmosphere. We used the Infrared Telescope Facility (IRTF)
Cryogenic Echelle Spectrograph (CSHELL), with a SBRC 256 x 256 InSb
detector. We obtained ~60 long slit spectra on and around Io with a
resolving power of ~18000, and a S/N of ~ 200. In addition, we searched
for [SI] emission from Io in eclipse by Jupiter using both imaging and
spectroscopic modes.

Sulfur and oxygen are the dominant atomic species in Io's atmosphere,
but are relatively difficult to observe. Ultraviolet emission from
atmospheric OI and SI has been observed from space with IUE and HST.
[OI] emission at 6300 angstrom has been monitored (Scherb and Smyth,
JGR, Vol. 98, p. 18729, 1993), but no corresponding visible [SI] line
exists. Detection of [SI] at 10825 angstrom would be the first
ground-based observation of neutral sulfur emission.

The goals of this study are to (1) detect [SI] emission (or find an
upper limit) and map it spatially about Io, to (2) determine the neutral
sulfur abundance in Io's atmosphere, and to (3) determine an IR/UV
brightness ratio of [SI] when matched with HST ultra-violet observations
(Clarke et. al, JGR, Vol. 99, p. 8387, 1993).

NASA grants NAGW-2484, NAGW-2492 and NSF PYI grant 9157751 have
supported this research.

20.17-P Wong M. C. Johnson R. E.
The Effect of Plasma Heating on Sublimation-driven Flow in Io's

The atmospheric flow on Io is numerically computed in three dimensions
for a flat geometry for a sublimation atmosphere on the trailing
hemisphere subjected to plasma bombardment, UV heating, and IR cooling.
Calculations are performed for subsolar vapour pressures of ~ 6.5 x
10^-3Pa (~3 x 10^18SO2/cm^2) and 6.8 x 10^-4Pa (~4 x 10^17SO2/cm^2), the
latter approximates the vapor pressure of Fanale et al. (1982). The
amount of plasma energy deposited in the atmosphere is 20% of the plasma
flow energy due to corotation (Linker et al., 1988). It is found that
plasma heating significantly inflates the upper atmosphere, increasing
both the exobase altitude and the amount of surface covered by more than
an exospheric column of gas. This in turn controls the supply of the Io
plasma torus (McGrath and Johnson, 1987). The horizontal flow of mass
and energy is also important in determining the exobase altitude; and it
is shown that IR cooling can be important, although our use of the
equilibrium, cool-to-space approximation for a pure SO2 gas (Lellouch et
al., 1992) may overestimate this effect. The calculated exobase
altitudes are somewhat lower than those suggested by McGrath and Johnson
(1987) for supplying the torus, indicating the details of the plasma
energy deposition and sputter ejection rate near the exobase, as well as
the IR emission from this region need to be examined. In addition, the
molecules sublimed (or sputtered) from the surface are transported to
the exobase in times short compared to the molecular photo-dissociation
time. Therefore, the exobase is dominated by molecular species and the
exobase is supplied by a small region of the surface.

This work was supported by NASA Planetary Atmospheres Program under
contract NAGW-461 and NSF Astronomy Division under contract AST-9120078.

20.18-P Thomas N.
Ion Temperatures in the Io Plasma Torus

Observations of the radial dependence of T(sub)perpendicular to of the
four major ion species in the Io plasma torus are presented.
Doppler-resolved forbidden line emissions of S^+, S^++, O^+, and O^++
were observed within a period of 72 hours on 3-5 March 1993 with the
3.6m telescope of the European Southern Observatory. The long slit
spectra of S^+ show an increase in T(sub)perpendicular to at the
"ribbon". Enhanced electron densities and [SII] emission were also
observed at this point. The observations are broadly consistent with
models of line profiles based on Voyager 1 PLS data. The signal to noise
for the S+ observations was sufficient to reveal two velocity components
in the distribution, at 33 and 188 eV, inside Io's orbit. Modelling
shows that the two components arise from the superposition of hot outer
ions upon a core of cool inner torus ions as a result of line of sight
effects. Outside Io's orbit, the temperature of S^+ ions was observed to
decrease from 120 eV at 6.0 R(sub)J to 80 eV at 6.8 R(sub)J consistent
with a cooling expansion of the plasma into the middle magnetosphere.

20.19-P Woodward R. C. Smyth W. H.
Modelling the Io Plasma Torus: Effects of Global Parameters

A semi-empirical model of the Io plasma torus is under development to
facilitate the interpretation of data from the Galileo spacecraft and
from ground-based and Earth-orbiting instruments. This model generates a
three-dimensional description of the torus species, from which simulated
spatial images and spectra of emission lines are produced for comparison
with actual data. Great care is being taken to allow maximum flexibility
in varying the physical parameters of interest, in order to determine
their effects on the generated images, and hence the restraints on these
parameters that the data imply. The structure and operation of the model
will be briefly described. We will then present a collection of images
generated by the model--both stills and animated sequences--showing the
global morphological changes produced by various distributions of torus
species and sets of physical conditions. In particular, the effects of
different magnetic field models, of System III (longitudinal)
asymmetries in the torus, and of a convection east-west electric field,
will be presented. This work is supported by NASA contract #NAS7-918.

20.20-P Schneider N. M.
Longitudinal Variations in the Io Torus

A comprehensive analysis of groundbased S+(6731 Angstroms) torus images
has revealed important insights on the torus' longitudinal variations.
Virtually all optical emission studies have reported significant
intensity variations with Jupiters System III longitude
(Lambda(sub)III), but the exact cause of the variations has not been
identified. Our observations confirm large intensity variations
throughout the torus, and identify the prominent 'ribbon' (just inside
Io's orbit) as the most variable region.

Further analysis shows that most plasma properties of the torus vary
with Lambda(sub)III, not just the intensity. The parallel ion
temperature (measured from the ribbon scale height) varies by at least a
factor of four, and is often anticorrelated with intensity. The radial
location and the latitude of the ribbon both deviate significantly from
the dipole approximation. This assortment of variations is important for
two reasons. (1) Different observational methods may record quite
different Lambda(sub)III intensity behaviors, as the multiple variations
amplify or cancel each other. For example, images (that pinpoint the
brightest equatorial emission) will show a different intensity variation
than tall slit spectra, (which depend both on equatorial intensity and
scale height). In addition, observations that do not track the complex
motion of the torus introduce spurious variations due to pointing
errors. (2) More importantly, these linked variations may hold the key
to identifying the underlying cause of the intensity variations.

We will discuss our results in the context of the numerous prior torus
studies at many wavelengths. We will also emphasize the important
ramifications that these multiple variabilities have for the torus and
Io's atmosphere.

NASA grants NAGW-2484, NAGW-2492, and NSF PYI grant 9157751 have
supported this research.

20.21-P Herbert F. Sandel B. R.
Tomographic Analysis of Voyager UVS Observations of the Io
Plasma Torus

Extensive observations of the Io plasma torus (IPT) by the Voyager 1
Ultraviolet Spectrometer (UVS) have been combined by tomographic
analysis. Interpretation of these data sets has in the past been
difficult because of the complicated geometry and changing aspect of the
IPT. The data have been fit by model spectra computed from parameterized
model spatial distributions of S^+ , S^++ , S^+++ , O^+ , and O^++ using
singular value decomposition (SVD) analysis.

A well-known difficulty with analysis of UVS IPT spectra has been the
mutual near-linear-dependence of the low-resolution spectra of S^++ ,
O^+ , and O^++. SVD analysis has been used to cope with this problem by
adjusting the combination of spectral eigenvectors to reflect the very
low O^++ /O^+ ratio known from ground-based observations to prevail at
magnetic L values near 6. Despite our constraining the O^++ /O^+ ratio
to ~0 at L = 6, the derived value rises at larger L, as found earlier by
Bagenal et al (GRL 19, 79, `92).

We find that the ionization level (including the O^++ /O^+ ratio) rises
with increasing L. In addition, there is little evidence in our data
that ion scale heights increase with L, suggesting that the parallel ion
temperature may not increase with L. If ion scale heights are constant
or decrease with increasing L, the extrapolation of Voyager in situ
plasma measurements in the outer IPT to the centrifugal equator would
need modification.

This work is supported by NASA grants NAGW 3440, 1498 and 3657.

20.22-P Matson D. L. Blaney D. L. Johnson T. V. Veeder G. J.
Goguen J. D.
The Temporal Evolution of Eruptions on Io

The size and temperature history of an eruption on Io can be
characterized from multi-wavelength radiometry. An extremely useful tool
for analyzing these data is what we call a Thermal Infrared Differential
Emission (TIDE) plot. The emission during an outburst (a period of
anomalously high 4.8 micrometers flux) is subtracted from the previous
base emission level to produce values for the excess spectral emittance
due to the eruption for each wavelength. Pairs of these flux differences
are then plotted for each observation during the night (e.g., 4.8
micrometers flux excess vs 8.7 micrometers flux excess). For a single
source model, each point on the TIDE plot defines a temperature and
effective radius for the model source. Lines of constant temperature and
radius can be overlain on the TIDE plot so that the trends in the data
such as cooling or increasing area with time are obvious. Examination of
the plot quickly reveals if a single source model is physically
reasonable for a given event. The graphical solution for temperature and
area can be seen immediately. This technique is applicable in modeling
both lava flows and lakes. We estimate the errors in the plotted
differences from the observed scatter of the non-outburst measurement
and use the TIDE plot to estimate the resulting errors in modeled
temperatures and area as a function of time.

We have applied this technique to the 1986 and 1990 outburst data. From
this we have derived the eruption histories of the temperature, area,
and power for these events.

This work was done at JPL/ CALTECH, under contract to NASA.
Also all authors were visiting scientists at the NASA Infrared Telescope

20.23-P Smyth W. H. Marconi M. L.
Iogenic Plasma Sources for Jupiter's Magnetosphere

Neutral cloud models that calculate the three-dimensional spatial
distribution of Na, K, O, S, and SO2 in the satellite coronae, extended
atmospheres, and neutral clouds have been developed and are being
applied to observational data for the neutrals to determine their gas
source conditions at the satellite exobase and to assess their impact on
the plasma torus. Through electron impact and charge exchange processes,
the various neutral species produce Iogenic plasma sources for Jupiter's
magnetosphere. The Iogenic sources give rise to ion loading, mass
loading, pick-up electrical conductivity, and an energy input, which are
important in determining the composition and structure of the plasma
torus and its outward transport dynamics in the larger planetary
magnetosphere. The spatial structure and density of the neutral clouds
and their Iogenic plasma source rates in the circumplanetary space
depend upon the absolute gas source rate at the satellite, the initial
velocity distribution for this source, (which determines the ensuing
orbital dynamics of the atoms and molecules), and the neutral sink,
which is dominated in the vicinity of Io's orbit by space-time dependent
electron-impact ionization of atoms (or dissociation of molecules) and
charge-exchange processes in the plasma torus. The instantaneous Iogenic
source rates therefore have a two-dimensional distribution relative to
Io's location [i.e., radial (L shell) and longitudinal]. This
two-dimensional distribution also changes. It depends upon two
additional angular quantities that specify Io's position in the plasma
torus and about Jupiter: the Io System III longitude angle and the Io
phase angle. The latter angle is needed because of east-west asymmetries
in the plasma torus. Calculations for the space-time dependence of the
Iogenic plasma sources produced by System III longitude will be
presented for Io both east and west of Jupiter. Progress in
understanding the implications of these Iogenic distributions on the
plasma torus will be discussed. This work is supported by NASA contracts
NASW-4804 and NAS7-918.

SESSION 21 ....... Rings I
Thursday, 8:30 - 10:00 Crystal Ballroom A
N. Rappaport and C. D. Murray, Moderators

21.01 Dilley J.* Crawford D.
Size/Mass Dependence of Elasticity in Collisions of Icy Spheres

In an earlier work (J. Dilley, Icarus 105, 225, 1993), a viscous
dissipation model for the collisions of icy spheres was developed. This
model successfully described the velocity dependence found by Bridges et
al. (1984), and both the velocity and size dependence found by Hatzes et
al. (1988) for such collisions. On the basis of these results, it was
suggested that the coefficient of restitutions for collisions of
physical ice balls may well have a significant size/mass dependence.

We have recently tested this idea by having ice balls of four different
masses, ranging from 34 gm to 565 gm, collide with an ice block at a
speed of about 1 cm/s. We find that elasticity increases systematically
with size/mass, collisions becoming successively more elastic as mass
increases. The effect is not small: the mean coefficient of restitution
for the largest spheres is more than twice that of the smallest ones.
This finding may have important implications for ring systems, in which
the mass of individual particles may vary by many orders of magnitude.

This work was supported by a grant from the Ohio University Research

21.02 Mosqueira I.*
Local Simulations of Perturbed Rings

We investigate the role of viscosity on the dynamics of rings of optical
depth tau ~ 1 using a modified N-body code written by S. J. Aarseth in
which each particle is followed with its own time step. The modified
version includes periodic boundary conditions appropriate for a
perturbed shear flow, finite particle size collisions with a
velocity-dependent coefficient of restitution, and epicyclic motion in
the absence of particle interactions. In particular, we extend the
ring-patch approach of Wisdom and Tremaine (1988) Astron. J., 95, 925,
to cover rings with q = a de/da not equal 0 where a is the semimajor
axis and e is the eccentricity of the particle streamlines. In the
unperturbed limit (q = 0) we test our code against the results of Wisdom
and Tremaine in the absence of particle self-gravity, and against the
results of Richardson (1993, preprint) which include particle
self-gravity. Even though we have retained the option of running our
code with N-body particle self-gravity, our main results have been
obtained in the presence of an azimuthally local self-consistent mean
field. As applications of our code we have used observationally
constrained parameters for the eplison and delta Uranian rings, as well
as for the outer portion of Saturn's B ring. For our model parameters we
find that the assumption used in the self-gravity model that viscosity
does not affect differential precession (Goldreich and Tremaine (1979)
Astron. J., 84, 1638) is self-consistent unless steep fronts form for
sufficiently large q > 0.8. We find that understanding the effects of
viscosity for the Uranian rings will require that both local and
nonlocal effects be considered if the coefficient of restitution
experimentally obtained by Bridges, Hatzes, and Lin (1984) Nature, 309,
333, is appropriate for ring particles. We also find evidence that the
criterion for viscous overstability is satisfied in the case of high
optical depth rings, as originally proposed by Borderies, Goldreich, and
Tremaine (1985) Icarus, 63, 406. Unlike their analytical treatment,
however, our numerical results indicate that this criterion is satisfied
even in the presence of observable perturbations with q ~ 0.5 for an
average optical depth < tau >~ 2, making viscous overstability the
leading candidate mechanism to explain the nonaxisymmetric structure
present in the outer portion of Saturn's B ring.

21.03 Greenberg R.* Petit J. M.
Viscosity in Keplerian Disks: Steady-State Velocity
Distribution and Non-Local Collisional Effects

Angular momentum is transported in planetary rings by both local and
non-local effects. "Local" transport is due to particles crossing a
given boundary, preferentially transporting momentum from one side to
the other. "Non-local" transport is due to interactions between
particles that are not at the same location through a distant
interaction such as collisions between finite size particles or
gravitation. Our model of momentum transport is based on the description
of the physics by Greenberg (1988, Icarus 75, 527) and is structured to
allow non-local momentum transport due to collisions or collisions with
mutual gravitation between particles in cases where Hill's approximation
is valid. In either case, the particles are solid spheres, and
collisions occur with a coefficient of restitution epsilon for the
radial component of the relative velocity, but no tangential friction
(i.e. relative tangential velocity at contact point is unchanged).
Calculations reported here are two-dimensional and ignore the mutual

For the local viscosity, we find approximately the same optical depth
(tau) dependence as predicted by Goldreich and Tremaine (1978, Icarus
34, 227). For tau = 1, if \epsilon > 0.5 the velocity will grow
indefinitely. If \epsilon < 0.4, our results show velocity approaching a
non-zero velocity governed by particle size. The equilibrium value of
epsilon thus appears to be somewhat lower than calculated by Goldreich
and Tremaine.

We also study the evolution of the mean random velocity V assuming a
plausible relatioh between epsilon and V. The non-local viscosity
increases monotonicly with tau, unlike the local component. For cm-size
particles with tau = 0.1 the non-local viscosity is of the same order as
the local part (~10^-6 m^2/s), but for m-size particles, it is more than
100 times larger than the local one, because the random component of the
motion is comparable to the size of the particles. Thus viscosities
might be quite different from what has generally been assumed in past
modeling of ring systems. Continuing studies involve the effects of
mutual gravitation and of different sizes among the particles.

21.04 Stewart G. R.*
Collisionless Damping of Nonlinear Spiral Density Waves

A new mechanism is described for the damping of nonlinear spiral density
waves in planetary rings. Previous models of nonlinear spiral density
waves have focused on the damping caused by dissipative collisions
between ring particles (Shu, Dones, Lissauer, Yuan, and Cuzzi, 1985;
Borderies, Goldreich, and Tremaine, 1986). The collisional models found
a slow monotonic decrease in wave amplitude whereas the wave trains
observed in Saturn's rings typically display rapid modulation and
non-monotonic decay of wave amplitudes. Particle collisions cannot be
the cause of such rapid wave damping because the large optical depths
observed in the wave crests imply short mean free paths and thus a small
effective viscosity.

Rapid damping of spiral density waves can occur if one allows for the
possibility of nonlinear wave-wave interactions due to self-gravity.
Specifically, a wavetrain launched from a resonance location with a
particular azimuthal wave number, m, will tend to excite daughter waves
with wave numbers equal to 2m and the higher-order harmonics. The
original wavetrain is damped as it transfers angular momentum to the
higher-order harmonics of the total nonlinear wave spectrum. The damping
is not monotonic because some of the angular momentum can be transferred
back to the original parent mode with wave number m.

The theory of this collisionless damping is based on a phase-space
variational principle for nonlinear spiral density waves. Similar
variational principles have recently been derived to describe nonlinear
wave interactions in plasma physics (Crawford, Johnson, Kaufman, and
Oberman, 1986). If one constrains the variational principle to allow
only one azimuthal wave number, m, then the Shu, Yuan, and Lissauer
(1985) theory of nonlinear density waves is recovered. More general
solutions describing the excitation of higher-order m-harmonics are also
readily obtained from the variational principle. Detailed comparisons
with observed wavetrains are in progress.

This work was supported by NASA grant NAGW-2895.

21.05 Rappaport N.* Longaretti P. Y.
Simulation of the Epsilon Ring of Uranus

This talk deals with the problem of the origin of the rigid precession
of eccentric ringlets. One of the problems was discovered by Marouf,
Gresh and Tyler in 1977: The radio optical depth profiles of the epsilon
ring are inconsistent with the self-gravity model.

In previous work, we investigated the viscous instabilities that arise
in a dense ring where the stress tensor exhibits an anomalous behavior.
We showed that the librations of the streamlines are not completely
damped. This is a consequence of the fact that self-gravity imposes
non-constant q across the ring.

We performed numerical simulations of the epsilon ring in order to
assess effects of librations in the evaluation of the self-gravity
model. Some of the questions addressed by this work are: How do
librations modify the relation between the mass and the geometry of an
eccentric ringlet? How do librations affect the difference of periapsis
angles between the inner and outer edges of an eccentric ringlet? What
is the effect of changing the ratio between the self-gravity and viscous

21.06 Franzke P.* Spahn F.
Stability Analysis of a Planetary Ring

In the framework of hydrodynamics we have analized the stability of a
disk of dust taking into account the physical effects: Gravitation of
the central body, self-gravitation of the disk, pressure, and the
influence of the space dependent viscous transport coefficients. By
means of a linear stability analysis we have derived and solved a
dispersion relation for the case of planetary rings. The influence of
the physical effects mentioned above has been discussed separately.
According this analysis we have found that a planetary ring is stable
against small perturbations at scales L > 200m for moderate optical
depths tau < 1 ...1.5. However, in the range of scales 50m < L < 200m a
infinitesimally thin dust-disk with an optical depth tau > 1 ...1.5
becomes unstable caused by the interplay of viscous effects with the
self-gravity of the ring. In the isothermal regime-velocity dispersion c
does not depend on the optical depth tau - the pressure p is variable
propto c^2 of the particle ensemble stabilizes the disk at scales L <
50m. In this case, viscous effects -- inclusive their spatial variation
(variation with the optical depth tau) -- play only a minor role
(effects of higher order). However, laboratory experiments have shown
the temperature T propto c^2 to be a decreasing function of the density
sigma propto tau, which results in a less importance of the
stabilization due to pressure at small scales L < 50m, and which seems
to increase the importance of the viscosity v. Also the amount of energy
and momentum carried by acoustic waves inside the ring-particles depends
sensitively on the properties of particles' material. If one assumes
hard spheres, then only self-gravity is able to cause instabilities at
scales ~ 100m, as suggested by Fridman and Polyachenko (1984) and
numerically confirmed by Schmit and Tscharnuter (1994). If one takes
into account a damping of these acoustic waves the viscous effects
become sensibly more important and shift the unstable region to smaller
optical depths. This could be the case if the ring-particles does not
consist of purely cristalline ice but more or less of loose aggregates
of water frost mixed with dust.

21.07 Espresate J.* Lissauer J. J.
Dynamics of Test Particles Near a 2:1 Orbital Resonance

We investigate the motion of massless (test) particles orbiting a planet
in the presence of a small satellite. A systematic study is performed on
the behaviour of the osculating orbital elements of test particles as a
function of time for a variety of initial conditions. In particular, we
include a range of initial particle semimajor axes surrounding exact
resonance. We consider satellites on circular and eccentric orbits. We
use planetary gravitational potentials which are either keplerian or are
modified by a small oblatenes term. As part of our investigation, we
compute the transient torque of the satellite on a disk of particles
having initially circular orbits and the region in semimajor axis over
which the torque is exerted.

This research was supported in part by NASA's Planetary Geology and
Geophysics grant NAGW-1107 and by DGAPA at Universidad Nacional Autonoma
de Mexico.

21.08 Murray C. D.*
Estimating the Lifetime of the Rings of Saturn From
Observations of the Orbit of Prometheus

One of the most important questions in ring dynamics concerns the
lifetimes of the ring systems of the outer planets. In the case of
Saturn the main ring system should be collapsing due to the exchange of
angular momentum between the ring particles and the small satellites
which produce density waves at resonant locations in the rings. The rate
at which the satellites move outwards determines the lifetime of the
rings. Accurate observations of the evolution of the orbit of
Prometheus, the inner F ring shepherding satellite, would lead to direct
measurements of this rate. On theoretical grounds Lissauer et a1.
(Icarus 64, 425-434, 1985) estimate that the change in Prometheus' mean
motion is (-1.3 x 10 ^-8)degrees d^-2 although the rate may be even
slower if Prometheus' density, like that of,Janus and Epimetheus, is as
low as ~ 0.6 g cm^-3 (Nicholson et al., Icarus lO0. 464-484, 1992). The
problems associated with rate measurements and the need for continuous
observations over long time periods are discussed. Images from the
Cassini spacecraft, due to go into orbit around Saturn in 2004, will
provide the best means of determining the lifetime of the ring system.

This work was partly funded by the UK Particle Physics and Astronomy
Research Council.

21.09 Srama R.* Grun E. Fechtig H. Jessberger E. K. Pernicka E.
Lura F. Mohlmann D. Wasch R. Ahrens J. Auer S.
Cruise A. M. Havnes O. Igenbergs E. Johnson T. V. Lamy P.
Morfill G. E. Schwehm G. H. Svestka J. Tuzzolino A. J.
Capabilities of the Cosmic Dust Analyzer (CDA) for the Mission

One major question in space science is the origin of small dust
particles in our planetary system. The CDA is a german experiment which
will be flown on the spacecraft CASSINI launched on October 1997. The
scientific goal is the exploration of the Saturnian ring system, the icy
satellites and the magnetosphere of Saturn, the Jovian environment and
the properties of interplanetary dust. The CDA consists of two
subinstruments: The Dust Analyser (DA), caused by his sensitive area of
0.1 m^2, dust fluxes as low as 1 impact/month can be detected, and the
High Rate Detector (HRD), which is used to measure higher fluxes up to
1000 impacts/s (Saturnian ring plane crossing). The physical processes
used to detect the dust, are impact ionisation and the depolarisation of
a foil, respectively. the CDA has the capability to measure the velocity
(1-30 km/s), the charge (10^-15 _10^-12 C), the mass (10^-15 _10^-9 g)
and the chemical composition of particles in micrometers size. A
laboratory model of the instrument is complete and under test. Data from
calibration measurements at the dust accelerator in Heidelberg are

21.10-P Hertzsch J.-M. Scholl H. Spahn F.
Simulations of Collisions in Planetary Rings

We present results of 2- and 3-dimensional high-resolution simulations
with 2^19 particles of the formation of structures -- ringlet, gaps, and
wakes -- in a planetary ring due to an embedded satellite. The
calculations were performed on a Connection Machine CM-200 and on a CRAY
YMP-EL. In previous work, where we did not take into account collisions
between the ring particles, we have already been able to follow the
evolution of these patterns. Now we have included in our model inelastic
collisions in order to investigate their influence on the formation of
structures. The collisions were treated in a rigorous way using the
velocity dependence of the restitution coefficients. Here we have
applied results of theoretical investigations^1 and of laboratory
experiments^2 on the energy loss of colliding ice particles.

We have investigated the persistence of the structures for a collision
probability derived from the observed optical depth for Saturn's ring
near the Encke division the mentioned structures -- ringlet, gaps, and
wakes - remain clearly visible over at least 100 revolutions. ^1
Brilliantov et al. 1994, to be published. ^2 Hatzes et al.1988, Mon.
Not. R. Astr. Soc. 231, 1091 -1115

21.11-P El-Deeb M. Marouf E. A.
Time-Frequency Representation of Wave-Like Features in Saturn's

The time-frequency representation depicts the variation of the spectral
content of a signal with time. It is particularly suited for
characterization of the non-stationary nature of Voyager ring
occultation observations. In the case of radio occultation, for example,
inference of ring particle sizes requires tracking in time and frequency
the contribution of individual ring features to the observed scattered
signal. Optimization of achievable spatial resolution requires
optimization of the tiling of the time-frequency plane to take advantage
of observed near-linear time-frequency drift of resolved features. In
the case of both radio and stellar occultations, many observed ring
features (in particular, resonant waves and embedded satellite wakes)
exhibit significant change in local spatial scale (frequency) with
radial distance (time). The time-frequency representation of these
features is ideal for identification of the nature of their dispersive
behavior. Here, we limit our attention to dispersive analysis of
selected wave-like features in Saturn's rings. Traditionally, the
short-time Fourier transform (STFT) provides the main tool for
time-frequency analysis. To overcome the limited frequency resolution of
the STFT, we analyze comparatively the performance of four other
approaches: (1) the (modified) Wigner-Ville distribution, (2) AR
modeling of the nonstationary signal, (3) AR modeling of the
Wigner-Ville distribution, and (4) the wavelet transform. We also
investigate algorithm performance for synthetic signals under various
noise conditions to characterize the systematic and random
uncertainties. Either the amplitude or phase of the time-frequency
representation is used to estimate variation of the instantaneous signal
frequency with time. Reliable estimation of the spatial frequency
profile of wave/wake ring features is immediately relevant to reliable
determination of ring physical properties and/or satellite orbit
parameters (v., e.g., Rosen et al., Icarus 93, 3, 1991; Showalter et
al., Icarus 66, 297, 1986).

21.12-P Jurac S. Johnson R. E. Morfill G. E.
Charging and the Sputtering Erosion of the E Ring Ice Particles

Horanyi et. al. (1992) calculated the motion of charged micron-sized
grains launched from Enceladus as a presumed source of the E ring
material. They showed that in the presence of a gravity field, solar
radiatlon pressure and electromagnetic forces grains could have a
spatial distribution which has many of the characteristic of the
observed E ring. Hamilton and Burns (1994) proposed that E ring could
sustain itself, i. e. the charged grains comprising the ring strike
Enceladus at high velocity ejecting new materials. Morfill et. al.
(1993) calculated grain potential of Saturns E ring particles for
various secondary electron emission parameters and proposed that the
sputtering rates of the E ring could be affected by the electrical
potential of the grains.

Utilizing the new laboratory data for secondary electron yields we
recalculate values of the E ring grain potential, which plays crucial
role in all mentioned models. We show that extrapolation procedures for
obtaining a grain potential used previously can not be done for the low
energy plasmas in the Saturnian and other planetary magnetospheres
having electron temperatures lower than 30 eV. For low energy plasmas,
not only secondary emitted electrons, but also elastically and
inelastically reflected electrons, constitute emitted currents from the
grain and significantly influence its equilibrium potential. In
addition, the threshold energy for secondary electron emission should be
taken into account whenever a significant portion of electron plasma is
below this threshold energy, i. e. does not produce secondary electrons.
We find E ring grain potential, varying from -6 V to +5 V at a radial
distances between 4 and 1O Saturnian radii, too small to influence ion
sputtering significantly. We calculate the sputtering rates and estimate
sputtering erosion of the E ring grains and compare this to the loss
processes proposed by Horanyi et. al. (1992).

REFERENCES Hamilton, D. P., J. A. Burns, Origin of Saturns E ring: Self
Sustained, Naturally, Science, 264, 550, 1994. Horanyi, M., J. A. Burns,
D. P. Hamilton, The Dynamics of Saturns E Ring Particles, Icarus, 97,
248, 1992. Johnson, R. E., Energetic Charged-Particle Interactions with
Atmospheres and Surfaces, Springer-Verlag, Berlin, 1990. Johnson, R. E.,
Rough stuff at Saturn, Nature, 363, 300, 1993. Johnson, R. E., D. E.
Grosjean, S. Jurac, and R. A. Baragiola Sputtering, Still the Dominant
Source of Plasma at Dione ?, Eos, AGU, Vol. 74, No. 18, November 30, p.
569, 1993. Morfill, G. E., O. Havens, C. K. Goertz, Origin and
Maintenance of the Oxygen Torus in Saturns Magnetosphere, J. Geophys.
Res., 98, 11285, 1993.

21.13-P Spahn F. Brilliantov N. K. Hertzsch J.-M. Poschel Th.
About Collisions Between Granular Particles: Application to
Planetary Rings

We propose a model for collisions between particles of granular
material. For impact velocities much smaller than the sound speed of the
particle-material we derive the restitution coefficients for normal and
tangential motion epsilon(sub)N and epsilon(sub)T taking into account
the dissipation in the collision dynamics as functions of the impact
velocity. For the normal restitution the Hertz-model has been extended
to dissipation and the results fit quite well to laboratory measurements
of Bridges et al.(1984). For the tangential restitution coefficient the
results reveal a rather nontrivial behaviour. Namely the type of the
collision (smooth sphere or rough sphere model) is determined by the
normal and tangential components of the impact velocity as well as by
the surface properties assumed of the granular particles.

Finally, we briefly discuss the results of our investigations with
respect to the kinetics of the particle ensemble in planetary rings. It
is pointed out that especially the particle's properties determine the
effectivity of local (random walk diffusion) and nonlocal (acoustic
waves inside the particles -- or hard spheres) transport of energy and
momentum. The latter prevents only viscous instabilities (Araki and
Tremaine 1986, Wisdom and Tremaine 1988) if this acoustic transport is
not strongly damped by ring-material itself. However, this can really be
the case due to cracks inside the ring-particles or even if the
particles consist more or less of conglomerates of ice, frost and dust
(dirty snowballs), where low values of the restitution-coefficients
related to a high amount of dissipated energy can be expected. Thus, we
conclude that the effect of viscous instabilities can really cause
structures or can prevent edges to spread in ring-regions where the
particle properties deviate considerably from the assumption of hard
Bridges et al., Nature 309, (1984), 333.
Wisdom and Tremaine Astron.J. 95 (1988), 925.


SESSION 22 ....... Venus
Thursday, 8:30 - 10:00 Crystal Ballroom B
F. P. Schloerb and T. Widemann, Moderators

22.01 Magalhaes J.* Seiff A.
A Search for Atmospheric Wave Activity in Pioneer Venus Entry
Probe Data Acquired Within and Below the Clouds

Observations of temperature and pressure from the atmospheric structure
experiment on the four Pioneer Venus entry probes have been analyzed for
evidence of small scale and periodic structure ln the lowest 70 km of
the atmosphere. Standard high pass digital filtering techniques were
used to isolate "eddy" structures with vertical scales less than about
10 km from a smooth "background." Observations from all four probes show
the eddy temperature (T') has an amplitude ~0.2-2 K and a vertical
wavelength of ~3-4 km over a wide range of altitudes. The vertical
velocity (w'), derived independently from pressure data using the
assumption of hydrostatic balance, shows amplitudes of ~1 ms^-1 and a
vertical wavelength of ~3-4 km at most altitudes from all four probes as
well. Over a portion of the altitude range, T' and w' are 90 degrees out
of phase as would be expected for a vertically propagating atmospheric
gravity wave. We are comparing these observations to 3-D linear gravity
wave theory. For this wave to be understood as a linear gravity wave,
the Doppler-shifted zonal phase speed must be comparable to or greater
than the wind speed variation over the region of propagation. This
constraint along with the dispersion relation of a 3-D linear gravity
wave imply the following wave properties: (1) the zonal wavelength of
the wave must be much greater than the vertical wavelength; (2) the
meridional wavelength must be much smaller than the zonal wavelength,
implying the observed wave must be a fully 3-D gravity wave with
constant phase lines in the east-west direction. Further analysis of the
wave interpretation as well as an evaluation of vertical fluxes of
energy and momentum will be presented. This work has been supported by
the Planetary Atmospheres Program of NASA.

22.02 Hinson D. P.* Jenkins J. M.
Magellan Measurements of Atmospheric Waves on Venus

Radio occultation data were acquired on three consecutive orbits of the
Magellan spacecraft in October 1991. Each occultation occurred over the
same topography (67 degrees N,127 degrees E) and at the same local time
(22 hours 5 minutes), but the data are sensitive to zonal variations
because the atmosphere rotates about 10 degrees per orbit. Comparisons
between observations and theory demonstrate that intensity
scintillations and small-scale temperature oscillations are caused by a
spectrum of internal gravity waves. Vertical wavelengths range from 30 m
to 3 km. We estimated the wave frequency, omega, by modeling the effect
of radiative attenuation and by fitting theoretical models to measured
spectra of intensity scintillations. Both methods yield omega ~ 2 x
10^-4 rad sec^-1; the ratio of horizontal-to-vertical wavelengths is
typically 50-100. The waves appear to be freely propagating, rather than
trapped, and omega << N, the buoyancy frequency, throughout the vertical
range of the observations. This includes the middle cloud layer, which
does not appear to be convective. There is a strong similarity between
intensity scintillations observed here and previous measurements, which
pertain to a wide range of locations and experiment dates. This implies
that the same basic phenomenon underlies all the observations and that
gravity waves are a persistent, global feature of Venus's atmosphere.
The most prominent gravity wave appears above the middle cloud. The
amplitude and vertical wavelength are about 4 K and 2.5 km,
respectively, at 65 km. The wave is nearly stationary relative to the
surface, which suggests topographic forcing as the source. Radiative
attenuation becomes appreciable at altitudes above ~65 km. This leads to
wave drag on the mean zonal winds of about +1 m sec^-1 per day
(eastward), which may contribute to the decay of the zonal winds that
has been observed previously at these altitudes. Temperature
oscillations with larger vertical wavelengths were also observed on all
three orbits, but we are unable to interpret these unambiguously at
present. This work is supported by the NASA Venus Data Analysis Program.

22.03 Buhl D.* Goldstein J. J. Chin G.
Observations of Variations in the Mesospheric Winds of Venus

The atmosphere of Venus contains a large amount of CO in the upper
mesosphere. The abundance at 100 km altitude is about 1000ppm. The
millimeter rotational lines of CO at 115 and 230 GHz have been used to
study the abundance and time variations of the Venus upper atmosphere
(Clancy and Muhleman, Icarus 64, 157; 64, 183; 89, 129). At the center
of the CO line is a narrow Doppler core which can be used to measure the
velocity of the mesospheric winds (Buhl, Chin and Goldstein, ApJ. 369,
L17). The millimeter CO lines arise from the 100 km altitude in the
Venus mesosphere. In modeling the mesospheric winds, the line of sight
velocity components across the disk are convolved with the telescope
beam and a vertical wind shear is included to account for a shift in the
altitude observed as the beam moves from disk center to the planet
limbs. Observations were made of this line during the inferior
conjunctions of Venus on January 1990 and Aprll 1993. Wind velocities of
approximately 100 m/s were measured and two distinct flows were
observed. Each observation showed a retrograde zonal wind flow of 100
m/s at the equator. The second component of the global wind is a
sub-solar to anti-solar flow driven by solar heating of the upper
atmosphere. This flow was first observed by Goldstein (1989) at
altitudes above ~100km. Our observations showed a return flow of gas
coming from the dark side of the planet during the pre-inferior
conjunction periods of December 1989 (120 m/s) and March 1993 (60 m/s).
For the post-inferior conjunction period of April 1993 the wind flow
reversed and we measured a direct flow of 140 m/s. These dramatic
changes in the sub-solar wind patterns suggest strong dynamical shifts
in the Venus mesosphere.

22.04 Zasova L.*
Structure of the Middle Atmosphere of Venus and of its Upper

Vertical and horizontal (latitude and solar related) atmospheric and
cloud structures retrieved by self consistent method from IR
spectrometry carried out on Venera-15 are discussed. In general, the
atmosphere is rather homogeneous at low latitudes (<55N) and very
inhomogeneous at higher latitudes. At high latitudes the temperature
differences at 100 mb level exceed 40 K and the upper boundary of clouds
(taken at frequency of maximal absorption of sulfuric acid) changes from
59 to 72 km/s. cold collar (60_80 N) and hot "dipole" (75_85 N) were
also observed. cold collar revealed itself as temperature feature,
positioned under 40 mb (~70 km) level. It shows the presence of several
local temperature minima which disappear at levels above 40 mb and
reappear again above 0.5 mb. The coldest areas were observed on the
morning side, and the warmest ones--on the night side of the planet
(near the evening terminator). Hot "dipole" was found to be a cloud
feature (the "hole") located at the altitude from 56 to 62 km. Its
origin is related to the downward meridional flux.

22.05 Schloerb F. P.* Jovell A. J. de Pater I. Grossman A.
The Limb Darkening of Venus at 2.9 mm Wavelength

A measurement of limb darkening may often be used to place valuable
constraints on the temperature and opacity structure of a planetary
atmosphere. In previous work by Good and Schloerb (Icarus 53 538, 1983),
a measurement of the limb darkening of Venus at 87 GHz was used to place
constraints on the abundance of SO(sub)2 near the base of the Venus
clouds. In this paper, we present new results of observations of the
limb darkening of Venus at three frequencies close to a strong SO(sub)2
absorption at 104 GHz (2.9 mm wavelength). The new measurements were
obtained during February 1994 using the Berkeley-Illinois-Maryland Array
(BIMA) in its B configuration. This 6-element array configuration
provided 15 interferometer baselines which well sample the visibility
function of Venus in the vicinity of its first null. The location of the
first null of the visibility function is very diagnostic of the amount
of limb darkening on a planet and provides a measurement that is free
from many types of systematic errors which plague normal interferometric
imaging. Thus, in situations which require high precision to constrain
or observe limb darkening effects, it is often most effective to measure
the null position and interpret the, results using atmospheric models.
Observations were made on two separate days, and they were arranged to
take advantage of the capability of the interferometer to measure
visibilities at two nearby frequencies simultaneously. This permitted an
accurate relative measurement of the amount of limb darkening at the
frequency of the strong S0(sub)2 absorption and at a second frequency
above or below this frequency. The relative amounts of limb darkening
observed inside and outside of the S0(sub)2 absorption region will then
be used to provide a measurement of the abundance profile of SO(sub)2
within the lower part of the Venus cloud deck.

22.06 Widemann T.* Bertaux J. L. Moroz V. I.
Vega-1 and 2 Descent Modules: In-situ Measurements of
Ultraviolet Absorption and Relationship of SO2 with Presently
Active Volcanism on Venus

In 1985, the Vega-1 and Vega-2 spacecrafts launched two probing missions
through the atmosphere of Venus. On their board, the ISAV ultraviolet
spectroscopy experiment generated a wealth of spectra in the 220-400 nm
range with an unprecedented vertical resolution (165 to 70 m).

The vertical profils of sulfur dioxyde yield a mixing ratio n (sub)so2
(z) = (37 +/- 4) ppm and n (sub)so2 (z) = (34 +/- 5) ppm at 22 km, and
~120 ppm betweem 30 and 40 km, a result confirmed by recent ground
infrared observations (Bezard et al., 1993). As thermochemical
equilibrium between SO(sub)2 and surface minerals at 735 K implies a
maximum value of only a few ppmv for SO(sub)2 (Fegley and Treinman,
1992), we discuss potential chemical sinks for S0(sub)2 between the
surface and z = 25 km, including OCS and H(sub)2S.

The bimodal distribution of H(sub)2S0(sub)4 droplets clearly shows
evidence for chemical contamination of mode 1, mainly upper a cloud
layer particles. UV spectral signature of the second absorber, found in
our data, is compared to available data for potential absorbers in the
planetary spherical albedo between 280 D and 380 nm. The relative size
of mode-l and mode-2 droplets constraints the partial vapor pressure of
gaseous sulfur species at A cloud altitude (45 - 70 km),
H(sub)2S0(sub)4, SO(sub)3 and S0(sub)2. We relate this new constraint to
n (sub)so2 (z) at cloud altitude. s

We finally discuss the implication of n (sub)so2 (z) for active
volcanism D and constraints on atmospheric dynamics and superrotation.

22.07 Fegley B. Jr.* Klingelhofer G. Lodders K.
The Formation of Hematite on the Surface of Venus

We present experimental results showing that at Venus surface
temperatures Fe^2+-bearing minerals (e.g., pyrite, pyrrhotite,
magnetite) are oxidized to hematite in CO2 gas mixtures having CO number
densities close to those (~2 times higher) at the surface of Venus. The
results, in combination with our basalt oxidation experiments (Fegley et
al. (1994) Meteoritical Society, Prague, Czech Republic) suggest that
the red color observed by Pieters et al. (1986) at the Venera 9 and 10
landing sites is due to sub-aerial oxidation of Fe^2+-bearing phases on
the surface of Venus, and that hematite, instead of magnetite, is
present on the surface of Venus. The experimental results provide key
constraints on chemical analyses, radar observations, and spectral
measurements from past Venus missions and are important for the design
of planned geochemical landers for future Venus exploration. The
experiments also demonstrate the great utility of Mossbauer (MB)
spectroscopy for providing crucial information about the mineralogy and
oxidation state of Fe-bearing phases on the surface of Venus from a
geochemical lander.

The experiments were conducted by isothermally heating well
characterized mineral samples and reagent grade chemicals in CO2 gas
mixtures with CO and SO2 at atmospheric pressure for periods of hours to
days. The reacted samples were examined by X-ray diffraction, MB
spectroscopy, and optical microscopy to determine their composition and
the extent of reaction. Typically, pyrite thermal decomposition
experiments gave samples consisting (from inside to outside) of pyrite
(if not completely reacted away), pyrrhotite, magnetite, maghemite, and
hematite. This was unexpected because magnetite (Fe^2+Fe^3+O4) and not
hematite (Fe2^3+03) is thermodynamically stable in the gas mixtures used
(and under Venus surface conditions). Also, experiments with synthetic
magnetites showed oxidation to hematite in CO-CO2 gas mixtures with
compositions inside the magnetite stability field. Based on these
experiments and others involving several solid state oxygen fugacity
"buffers" we propose that the formation of hematite occurs because the
CO and CO2 in the gas mixtures do not equilibrate with each other
(despite the presence of Pt thermocouples and wires in the hot zone).
The more abundant gas, CO2, controls the oxidation state of Fe-bearing
minerals in contact with it, thus leading to the formation of hematite
instead of magnetite.

Acknowledgments. The work at Wash. Univ. was supported by the NASA
Planetary Atmospheres Program (B. Fegley, P.I.).

22.08 Moore W. B.* Schubert G.
Crustal vs. Thermal Compensation of Venus Highlands

The variations in average depth of compensation (ADC) and
geoid/-topography ratio (GTR) between volcanic rises (e.g. Atla, Beta,
Bell, and Western Eistla Regiones) and other "plateau-shaped" highlands
(Ovda, Thetis, Tellus, and Alpha Regiones) on Venus have been attributed
to differences in compensation mechanisms. The large GTRs (20-35 m
km^-1) and deep ADCs (100-300 km) found for volcanic rises have been
interpreted as evidence for either dynamic or thermal support, while
compensation by crustal thickness variations is invoked to explain the
smaller GTRs (</- 10 m km^-1) and ADCs (< 100 km) of the plateau
highlands. Both thermal and crustal compensation predict that the geoid
height anomaly depends quadratically on the local topography. The
quadratic term depends inversely on the density contrast at the
compensating interface and both it and the density contrast will be
positive for crustal compensation and negative for thermal compensation.
From quadratic regressions of geoid vs. topography we consistently find
negative density contrasts at volcanic rises (~-3% at Beta, ~1% at Atla)
and either indeterminate (Alpha, Thetis, Tellus) or positive density
contrasts (~12% at Ovda) at plateau highlands, supporting the
interpretation that the former are thermally compansated at the base of
the lithosphere and the latter are compensated at the base of the crust.
Lithospheric thicknesses beneath volcanic rises are found to range from
150-300 km and the crustal thickness at Ovda Regio is ~70 km. These
thicknesses agree with previous ADC estimates. Our determinations of
density contrasts and modes of compensation are new.

22.09 Pettengill G. H.* Ford P. G. Simpson R. A.
Venus: Results from the MGN Bistatic Radar Experiments

The Magellan spacecraft, now in low circular orbit around Venus, carried
out bistatic radar observations of that planet in November, 1993, and
again in June, 1994. The spacecraft illuminated the surface using its X-
and S-band telemetry transmitters and high-gain antenna directed to
permit specular reflection as viewed from Earth. Echo signals were
received using dual circularly polarized receiving channels, in order to
permit a full Stokes-parameter description of the scattered signal.

Observations from the November experiment covered an arc over the Venus
surface moving from 60 degrees S, 75 degrees E up through the "horns" of
western Ovda Regio, continuing to about 77 degrees N, then over to 75
degrees N, 173 degrees E, a total strip length of more than 17,000 km.
During this interval, the angle of incidence to the surface varied
slowly from about 75 degrees to 80 degrees. The June experiment explored
a much shorter arc that traversed Maxwell Montes, and accomodated a
wider range of incidence angles.

In these experiments, the Doppler spectral width of the echo is
determined primarily by the "footprint" of the spacecraft's antenna,
although the spectrum occasionally narrows where a locally smooth region
on the surface dominates the scattering. Such an echo was observed at
5.1 degrees S, 66.5 degrees E, just to the east of Verdandi Corona. This
anomaly, smaller than 6 km in at least one dimension, appears to lie
within a radar-dark area, surrounded by tesserae that appear to have
been embayed by smooth flows. It is possible, but not probable, that it
may originate in a suitably smooth and inclined portion of the Verdandi
Corona, although this is difficult to establish with certainty.

This work was supported under JPL Contract 957070, as well as NASA Grant

22.10-P Fomin N. N. Yanovitskij E. G.
Absorption Line Formation in an Optically Thick, Weakly
Absorbing, Vertically Inhomogeneous Planetary Atmosphere

We study absorption line formation in an optically thick, vertically
inhomogeneous atmosphere. It is assumed that absorption is weak in both
the absorption line and continuum. The problem is rigorously solved for
the following two specific cases: (1) isothermal slab with pressure
varying accordance with the polytropic law, and (2) polytropic model of
the atmosphere (for the central line frequency). A method is suggested
for computing the line profile in the case of an arbitrary dependence of
the optical properties of the atmosphere on the optical depth. The
method is based on reducing the corresponding boundary value problem to
the Cauchy problem. This theory is applied to analyze the behavior of
weak CO(sub)2 absorption bands in the atmosphere of Venus.

22.11-P Dlugach J. M. Yanovitskij E. G.
On the Effect of Width and Orientation of Spectrograph Slit on
Observed Planetary Spectrum: An Application to the Venus

We consider a weak absorption band for an atmosphere of a large optical
thickness. Absorption is assumed to be weak (nearly conservative
scattering in both continuum and the absorption band. In this case, the
equivalent width of the band can be represented as W = V (alpha) Q,
where for a fixed phase angle alpha the value of V (alpha) is dependent
on both the optical properties of tl atmosphere and the width and
orientation of the spectrograph slit. We have derived appropriate
formulas for computing the function V(alpha). Computations have been
performed for the case of the Venus atmosphere and an infinitely narrow
and infinitely wide slit. The orientation of the slit was parallel and
perpendicular to the intensity equator of the planet. It is shown that
V(alpha) strongly depends on the width and orientation of the slit. We
have applied our computations to interpret spectroscopic measurements of
weak CO(sub)2 absorption bands performed by Young et al. (Acta Astron.
24, No. 1, 1974) We have found that the optical thickness of the Venus
atmosphere substantially decreases from the equator towards the poles.

22.12-P Yanovitskij E. G. Fomin N. N.
The Venus Cloud Layer: A Global Inhomogeneity?

A two-layer model of the Venus atmosphere is considered. It is assumed
that the pressure and temperature in both cloud and gas layers vary
according to the polytropic law. The observed oscillations of the
equivalent width of the R(0) line of the CO(sub)2 absorption bands at
788.3 and 868.9 nm are studied. It is shown that the observed equivalent
width variations can be explained in terms of a globally inhomogeneous
cloud layer. Specifically, we have found that the optical thickness of
the cloud layer is 34.4 for one hemisphere of the planet and 24.4 for
the opposite one. The ratio of the optical thicknesses coincides with
our previous independent estimate made on the basis of analyzing
variations of the integral brightness of Venus. The variations of the
height of the upper boundary of the cloud layer fully correlate with
observed variations of the degree of linear polarization in the UV
spectral region. The effect of the global asymmetry of the cloud layer
on the atmosphere superrotation is discussed.

22.13-P Roos Serote M. Coupe V. Drossart P. Lellouch E. Encrenaz T.
Saint-Pe O.
Infrared Imaging of Venus from IRTF/ProtoCAM Observations in

Observations of the nightside of Venus were performed from the 12th
through the 16th of October 1991 at the IRTF in Hawaii, using the
ProtoCAM instrument. A spectral resoluting power of about 60 was
obtained in the wavelength range of 3.6 to 5.0 micrometers. In this
range the upper clouds leave their signature in the form of a continuum,
with two CO2 bands at 4.3 micrometers and 4.8 micrometers superimposed.

The data consist of monochromatic images of about the half of the
nightside of Venus, with a spatial resolution of 0.35 arcseconds per
pixel, which corresponds to about 140 km per pixel. After calibration, a
spectrum at each point can be reconstructed. A rescaling with respect to
the Galileo/NIMS spectra was necessary because of incertainties in the
calibration and the measurements, but the rescaling factor was found to
be constant with wavelength, indicating that the results are in good
agreement with the NIMS-spectra.

Limb darkening measurements also show a good agreement with previous
results from the Pioneer- Venus and the Galileo missions. Finally, the
northern collar was cleary identified in the images in the 4.3
micrometers Co2 band, confirming the continuety of this feature.

22.14-P Ragent B. Yee S. Travis L. Crisp D.
Pioneer-Venus OCPP Imagery of Venus at 935nm Feature Velocities
and Correlations with Near Infrared and OCPP 365nm Imagery

Low contrast 935nm data taken by the Pioneer-Venus Orbiter Cloud
Photo-Polarimeter (OCPP) instrument, operating in its polarimetry mode,
have been processed and analyzed. In order to examine the small but
significant variations immersed in the background of scattered light.
two approaches have been attempted. The empirical "Minnaert" method.
previously used to process much of the OCPP 365nm imagery and a more
sophisticated technique involving an atmospheric model and multiple
scattering of the incident light have been used. Images using these
reduced data from a favorable period for such observations in June. 1988
have been prepared as projections on a Venus latitude-longitude grid.
and will be shown. More than 25 such images taken over eight consecutive
days in this period are available. Features present in these images have
been examined. and zonal velocity plots as a function of latitude.
obtained by measuring the displacement of these features in succeeding
images have been prepared. The velocities measured are typical of
atmospheric velocities previously measured in the middle cloud layer
region of the Venus atmosphere. Results of attempts to correlate the
features in these images with those in simultaneously obtained OCPP
imagery at 365nm. as well as with features in Earth-based near-infrared
imagery of the dark side of Venus taken during the same period will be
presented and implications considered.

B.R. and S.Y. acknowledge support under NASA Cooperative Research
Agreement NCC 2-466 and D.C. from the Planetary Atmospheres Discipline.
NASA Office of Space Science and Applications.

22.15-P Na C. Y.
Horizontal Variation of Venus SO2

The temporal variability of horizontal distribution of SO2 in the middle
atmosphere of Venus observed by Pioneer Venus are presented. The Pioneer
Venus UV Spectrometer (PVOUVS) obtained more than 500 multi-color images
of Venus from 1978 to 1992 which have been used to derive the global
average of SO2 amount. These multi-color images along with the spectral
data from the PVOUVS were used in discovering the long-term decline of
SO2 above the clouds of Venus (Esposito et al., 1988).

The multi-color images of Venus were obtained with various viewing
geometries, and thus these images provide the information on horizontal
distribution of SO2. The clouds of Venus are created by the
photochemical processes that oxidize upwelling SO2 (Esposito et al.
1979, Winick and Stewart 1980, Yung and DeMore 1982). Thus the
horizontal variation in SO2 distribution above the clouds of Venus may
be a manifestation of the cloud top dynamics. Del Genio and Rossow
(1990) have suggested that the cloud top dynamics of Venus may have a
cyclic variation with a time scale of 5-10 years. In addition, Clancy
and Muhleman (1991) showed a similar time scale for the change in CO
mixing profiles in the mesosphere of Venus. The time scale for the
changes in the horizontal distribution of SO2 will be presented, and it
will be compared to the time scales for other observed changes in the
atmosphere of Venus.

22.16-P Kliore A. J.
Changes in the Structure and Dynamics of the Venue Middle

The Pioneer Venus spacecraft was used to obtain 246 vertical profiles of
temperature and pressure in the neutral middle atmosphere of Venus from
December 1978 to January 1990. 114 of these profiles were measured
during the 1979-83 high solar activity period, with 39 dayside (SZA 90
degrees ) and 75 nightside measurements. During the 1984-86 low solar
activity period, 28 dayside and 35 nightside profiles were obtained, and
the 1989-91 solar activity period yielded 45 dayside and 24 nightside
observations. These measurements, which have uncertainties in
temperature of 15-20K at 90 km altitude, range down to about 45 km,
where the uncertainty is less than lK.

Significant variability was observed in the temperature structure at
high Southern and Northern latitudes and at the upper altitude range
(above about 50 mb of pressure). These variations could be temporal,
diurnal, or solar-cycle related. These temperature fields were used to
derive the cyclostrophic zonal winds for the various data sets
associated with solar cycle changes, and the resulting changes in the
persistent middle atmosphere circulation were derived. Although the
midlatitude jet is a persistent feature of Venus atmospheric
circulation, significant temporal changes were observed.

22.17-P Bougher S. W. Zhang S. Alexander M. J.
Role of Gravity Waves in the Venus Thermosphere

Wave like perturbations have been observed in the Venus upper mesosphere
and lower thermosphere by instruments aboard the Pioneer Venus Orbiter
and Probes (e.g. Kasprzak et al., GRL, 20, 2755, 1993). These
thermospheric gravity waves are thought to be launched from the Venus
cloud region, to propagate vertically, and to dissipate above 120 km due
to saturation. This "wave breaking" process provides significant
momentum and energy forcing to the thermosphere, thereby modifying the
large scale circulation from that otherwise expected from in-situ solar
forcing. In particular, the subsolar-to-ntisolar (SS-AS) symmetric winds
are much slower than expected from day-night pressure balance.
Furthermore, thermospheric density, temperature, and airglow data all
imply the presence of an asymmetric component of the global circulation
that is zonal and retrograde. Current theory holds that gravity wave
saturation and subsequent drag modifies this pressure balance, thereby
providing weakened SS-AS winds and an E-W asymmetry in the net wind
system (Alexander, GRL, 19, 2207, 1992). The NCAR Venus Thermospheric
General Circulation Model (VTGCM) has been modified previously to
incorporate energy and momentum forcing via the independent prescription
of Rayleigh friction, a superimposed retrograde zonal wind profile, and
eddy diffusion. These momentum flux and diffusion terms are in fact
coupled through gravity wave breaking theory. This paper describes a
self-consistent gravity wave breaking parameterization for use in the
VTGCM based on the terrestrial scheme of Fritts and Lu (JAS, 50, 3695,
1993). Adjustable wave parameters include: the phase speed of the waves
at their source, an E-W asymmetry factor, and the scale height for
growth of the wave energy density. Initial sensitivity tests are
presented which illustrate the relative importance of these parameters
on the simulated composition and airglow distributions, and the net wind

22.18-P Del Genio A. D. Zhou W.
GCM Simulations of Superrotation at Venus' Rotation Period

In recent years, it has been realized that the key to simulating
equatorial atmospheric superrotation in three-dimensional models of
slowly rotating planets lies in forcing the circulation with a
statically stable radiative heating profile associated with optically
thick high level clouds (cf. Del Genio et al., Icarus 101, 1-17, 1993).
But despite the fact that superrotation was first identified on Venus,
successful superrotation simulations have been carried out to date only
for rotation periods characteristic of Titan, which rotates more than an
order of magnitude faster. Limited modeling attempts at Venus' rotation
period yielded weak equatorial subrotation and jets of only 7-8 m/s
strength at high latitudes in the cloud layer. This conflicts with
simple Hadley cell-based superrotation theories, which predict that
superrotation should simply scale linearly between Titan's and Venus'
rotation rates if other parameters are fixed. Since all terms in the
momentum balance are quite small for Venus, computational accuracy may
be more important for the superrotation problem than for the simulation
of, e.g., the terrestrial general circulation. We will present new
simulations with a modified terrestrial GCM run at Venus' rotation
period that exhibit substantial equatorial superrotation (10-15 m/s) and
high latitude jets with speeds of 20-30 m/s in equilibrium, much closer
to the scaling predictions. The new model runs were conducted on an IBM
RISC6000 workstation, which performs computations more accurately than
the mainframe computer used for previous simulations; this change alone
more than doubles the globally integrated angular momentum achieved in
equilibrium with the model. The results were achieved starting from a
resting isothermal state, but runs initialized with a highly
superrotating circulation are also being conducted to test the
sensitivity of the equilibrium configuration to initial conditions.

22.19-P Hansell S. A. Wells W. K. Hunten D. M.
Optical Lightning on Venus

A search for optical lightning in the atmosphere of Venus has been
conducted. Imaging of the dark side of the planet was done on 5 separate
nights between February 25 and March 15, 1993. The imaging occurred at
the University of Arizona 61" telescope (Mt. Bigelow Station).
Coronagraphic optics were used to reduce scattered light from the bright
side of the planet. Over two hundred thousand images were obtained
during this period with a total integrated live time of three and one
half hours. Images were obtained with a frame transfer CCD running at a
transfer speed of up to 18.8 images per second. A narrow band filter
centered at 777.4 nm was used. This wavelength corresponds to the light
emission maximum for simulated lightning in a model Venus atmosphere.
(Borucki, W.J., et. al. (1983). Geophys Res. Lett. 10, 961).
Additionally, images were taken through a filter centered at 654 nm
(Hydrogen - alpha emission line) to be used as control.

The images have been reduced and several potential lightning flashes
have been seen. These flashes occur at a rate of about 5 per hour. There
is reason to believe that most of these are due to cosmic ray events.
However, there are 5 images which show signals highly indicative of
lightning flashes on Venus. The lack of similar events in the control
group supports this claim. These will be presented along with the
reduction method used. Additionally, approximations of the energy
associated with the flashes will be presented and compared to that of
Earth typical lightning. False detection rates and possible sources for
false detection as well as confidence in the detected flashes will also
be presented.

22.20-P Thompson T. W. Saunders R. S. Griffith D. G.
Magellan Flight Team
Magellan Mission Progress Report

The Magellan spacecraft has been operating in orbit around Venus since
August 10, 1990 with the objectives of (1) improving the knowledge of
the geological processes, surface properties and geologic history of
Venus by analysis of surface radar characteristics, topography and
morphology and (2) improving the knowledge of the geophysics of Venus by
analysis of Venusian gravity. The first objective was supported by radar
imaging, altimetric and radiometric mapping of the Venusian surface from
September 1990 until September 1992 (mission cycles 1,2,3). Some 98
percent of the surface has been mapped with radar resolution on the
order of 120 meters.

Radar data processing by JPL was completed in mid-1994. Over 1200 radar
image products are available as analog photographs and digital compact
disks (CD-ROMs) at the National Space Science Data Center (NSSDC),
Goddard Space Flight Center. In addition, the altimetric and radiometric
data, products from MIT and the cartographic maps from USGS are
available at the NSSDC. A contact for these Magellan data products is
the PDS Geosciences Node at Washington University, St. Louis, Missouri
(e-mail slavney@wunder.wustl.edu).

Some 950 orbits of high-resolution gravity observations were obtained
between September 1992 and May 1993 while Magellan was in an elliptical
orbit with a periapsis near 175 kilometers and an appoapsis near 8,000
kilometers. An additional 1000 orbits will likely be obtained since
orbit-circularization in mid-1993. These data exist as a 60 degree-by
60-degree harmonic field produced by Bill Sjogren of JPL. This and other
Magellan gravity products are available at NSSDC also.

Magellan has also provided information about the Venusian atmosphere.
The high effective power of the Magellan downlink radio system enabled
the deepest probing of the Venusian atmosphere by occultation. Also,
atmospheric drag measurements, particularly during aerobraking, have
expanded upon those originally obtained with the Pioneer-Venus
spacecraft and have validated the Venus International Reference
Atmosphere (VIRA).

22.21-P Tormanen T.
Morphological and Stratigraphic Characteristics of Ridge Belts
Associated with Tessera Boundaries on Venus

In latest survey of ridge belts associated with tessera boundaries [1,2]
63 ridge belts were identified on Venus from Magellan radar images. The
ridge belts were included in the survey based on their closeness to the
tessera margin and apparent influence of the tessera terrain on the
strike and location of the ridge belt. These belts have a total length
of about 25,000 km. Three types of ridge belts were identified based on
the strikes of the belts in relation to the tessera boundary and the
distance of the belt to the tessera margin: (1) Ridge belts directly in
contact with the tessera boundary; (2) Ridge belts not in contact with
the tessera margin but located less than 100 km from a large area of
tessera and clearly influenced by the presence of tessera terrain; (3)
Ridge belts striking in a large angle against the tessera margin and
terminating there [1,2].

Ridge belts associated with tessera margins fall into 4 morphological
classes based on ridge size, spacing, and morphology within a ridge
belt. Ridge belts of morphological class I consist of linear or arcuate
parallel ridges of even width. Ridge belts of morphological class II are
characterized by subparallel linear to sinuous ridges with varying
widths and lengths. Wide ridges are accompanied by narrower, more
sinuous ridges. Ridge widths typically vary along the belt. Ridge belts
of class III are composed of widely-spaced, arcuate, to slightly sinuous
ridges that do not usually branch. There are very few narrow ridges in
class III belts. Class IV ridge belts consist of wrinkle ridges with
usually a few wider and higher ridges. Ridges are typically sinuous and

Geomolphologic mapping of the ridge belts and their immediate
surroundings reveal that more than half of these ridge belts are not
embayed by adjacent plains deposits and they appear to have deformed the
plains that comes into contact and usually also embays the nearby
tessera boundary. There are also ridge belts associated with tessera
boundaries that deform earlier plains surface, which has since been
buried by later plains deposits and is only revealed at tbe ridge belt.
Occasionally, two ridge belts located in the same area (e.g., around
25N, 297E) display different stratigraphic ages: There appears to have
been two episodes of belt formation, which were separated by episode(s)
of plains emplacement. Only in few locations the ridge belts appear to
have deformed the tessera boundary itself. Examples of detailed mapping
and results of studies of relationships between ridge belt morphological
characteristics, tessera boundary type, and local/regional geology will
be presented and discussed.

References: [1] Tormanen (1993) LPSC XXIV, 1439. [2] Tormanen (1993)
Eos, 74, 379.

22.22-P You J. Kauhanen K. Raitala J.
Fractality of Ejecta Blankets of Craters on Venus

Fractal studies of outlines of crater ejecta found from radar data
allows some details studied, revealing information of impact craters and
certain environmental, geologic impact-related processes in the crater
formation. The role of various variables are seen from the Richardson
plots of the structured walk approach. A linear trend in the plot
indicates a fractal behavior and it is related to the lobateness of the
ejecta. Image resolution and edge-finding accuracy limit the use of the
method to scales ranging from 10 pixels to 1/10th of perimeter. The
first linear part below 10 pixels resembles that the non-fractal
Modified Koch Triad with missing high-resolution details.

Concluding from various plots the fractality increases with crater
diameter, ejecta at and perimeter length and is values related to impact
energy. With high impact energy values the ejecta material has flown
larger distances and covered higher areas with longer ejecta lobes. This
may be true also with outflows. Scattering within the fractal line part
indicates scales of the ejecta lobes or scales at which the used rulers
make shortcuts over them. The measure of the perimeter shortens
radically depending on how the outline winds or what is the actual scale
and location of lobes and bays of the ejecta blanke outline. The wider
the scattering is the higher the variation in such lobe dimensions is.
For every studied crater there is a maximun value beyond which the
fractality disappears. Its linear relation with various crater
dimensions is obvious because longer rulers can be used for larger
craters. When the ruler exceeds a threshold value it does not follow the
edge line resulting in a final scattering in Richardson plot. A few
common features were found in the Richardson plots. The scattering of
the second linear part be located close to its upper scale. The second
linear part may also consist of two smaller linear units or have a
remarkable deviation of data points in the middle. Finally, both first
and second linear parts may have the same trend.

22.23-P Baron J. Tyler G. L. Simpson R. A.
A Finite-Difference Time-Domain Study of Radiowave Scattering
from Buried Dielectric and Conducting Cylinders

Interpretation of radiowave scattering data from "rocky" planets such as
Venus has been limited by the lack of suitable models describing the
interaction of wavelength-scale inhomogeneities located on or buried
within the planetary regolith. We have developed a numerical method to
study the mutual coupling of buried dielectric and conducting objects,
in two and three dimensions, based upon the finite-difference
time-domain (FD-TD) technique. The basic methodology, which involves a
slight modification to the traditional FD-TD total-field/scattered-field
formulation, and preliminary results in two dimensions were described
last year (Wong et al., BAAS, 25, 1086,1993). Here we present (1) a
brief demonstration of the validity of the technique via comparison with
a (less general) moment method code, (2) a study of the scattered power
as the depth of a "rock" is varied within a lossless regolith, and (3) a
discussion of the coupling effects that arise when an ensemble of
scatterers is considered. The code is being generalized for three
dimensions; preliminary results from that work will be presented, if

SESSION 23 ....... Rings II
Thursday, 10:30 - 12:00 Crystal Ballroom A
D. P. Hamilton and M. I. Mishchenko, Moderators

23.01 Mishchenko M. I.*
Can Particle Nonsphericity Explain Radar Polarization Ratios
Observed for Saturn's Rings?

Saturn's rings exhibit unusually large linear (mu(sub)L = 1 0 +/- 0.3)
and circular (mu(sub)c = 0 35 / 1.25) radar polarization ratios at
centimeter wavelengths (Goldstein et al., Icarus 30, 104, 1977; Ostro et
al., Icarus 41, 381, 1980). Two physical mechanisms (or their
combination) can produce nonzero polarization ratios: multiple
scattering of radiation by a many-particle-thick medium and single
scattering by nonspherical particles. The multiple scattering mechanism
was discussed by Cuzzi and Pollack (Icarus 33, 233, 1978). On the other
hand, theoretical computations of light scattering by realistic
polydisperse models of nonspherical particles used to be extremely
complicated, and no attempt has been made to explore the possible
contribution of particle nonsphericity to the observed polarization

Recently, a very efficient method has been developed for computing light
scattering by polydisperse, randomly oriented, rotationally symmetric
nonspherical particles with sizes comparable to the wavelength of
radiation (Mishchenko, Appl. Opt. 32, 4652, 1993). We have used this
method in extensive nonspherical computations to model single-scattering
radar echoes for nonspherical ring particles. Specifically, we have
computed the linear and circular polarization ratios for a power law
distribution of randomly oriented oblate and prolate spheroids with the
index of refraction of water ice (1.78 + 0.001i). The range of effective
mean equivalent-sphere size parameters was [0, 20] and the range of
aspect ratios was [1.1, 2.2] for prolate spheroids and [1.1, 3.0] for
ablate spheroids. The computational results show that moderately
aspherical particles can produce large, wavelength-dependent circular
polarization ratios which can even exceed those observed for Saturn's
rings. Our computations also show that the spectral behavior of the
linear and circular polarization ratios for particles of a given size
and shape is fully correlated, thus indicating that they are produced by
the same optical mechanism. However, none of our spheroidal models has
produced a linear polarization ratio exceeding 0.6.

In summary, our computations show that single scattering by a shape
mixture of even moderately aspherical wavelength-sized ice particles can
explain the circular polarization ratios observed for Saturn's rings.
Thus, one does not need to invoke the multiple scattering mechanism and
require that Saturn's rings be many-particle thick. On the other hand,
the same particles cannot reproduce the large linear depolarization
ratio reported by Goldstein et al. (1977). As discussed by Ostro et al.
(1980), however, that number may be an overestimation of the actual
value due to difficulties associated with the observational technique
used. It should be noted that the multiple scattering mechanism also
encounters difficulties in explaining the reported number since weak
localization significantly lowers the linear polarization ratio
(Mishchenko, Earth Moon Planets 58, 127, 1992).

23.02 Marouf E. A.*
Stochastic Geometry Models for Diffraction by Planetary Rings

Stochastic geometry models provide a new powerful approach for
investigating extinction and near-forward scattering of electromagnetic
radiation by planetary rings. A discrete random medium composed of
particles of size large compared to the wavelength is replaced by a
randomly blocked diffraction screen placed in the plane immediately
behind the rings, as seen by the incident wave. Over the plane of the
screen the field assumes binary values: zero if the spatial location is
within a shadow area of a ring particle, the full-incident field
otherwise. For a given occultation geometry, the stochastic field
pattern is propagated to an observing receiver using the Huygens-Fresnel
principle. Statistical averages of the diffracted field follow from
statistical characterization of the stochastic geometry of the randomly
blocked screen. In particular, the first moment (the average area
fraction blocked) determines the ring optical depth, while the two-point
spatial correlation function determines the angular spectrum of the
near-forward scattered signal, or equivalently, the differential
cross-section per unit ring area. I consider first the special case of
normal incidence on a many-particle-thick (slab) ring model, assuming
spherical particles that can be spread in size. When projected on the
diffraction screen, the centers of the shadow circles are assumed to
form a two-dimensional Poisson point process. At each center, the
circular shadow area has a random radius determined by an assumed size
distribution; the shadow circles can overlap. Remarkably, analytical
results for the optical depth and differential cross-section obtained by
averaging over the stochastic screen geometry are in full agreement with
results previously derived using a classical radiative transfer approach
to the multiple scattering problem (Marouf et al., Icarus, 49, 161, 1982
and Icarus, 54, 189, 1983). The Poisson model is known to correspond to
a sparsely populated slab, hence to small volume fraction. More general
ring models for which closed form analytical solutions may not be
attainable can be numerically investigated using computer generated
realizations of the stochastic geometry of corresponding shadow areas.
The models may include arbitrary vertical ring profile, particle
crowding, particle clustering, and preferential cluster orientation.
Extension of the radiative transfer approach to such models is, at best,
extremely difficult. In contrast, the stochastic geometry model allows
porting of a host of well developed time-series and spectral analysis
techniques to characterize relevant observable statistical averages in
terms of physical ring model parameters.

23.03 Alix J.-M.* Festou M. C.
On the Presence of Subcentimeter Particles in the A and B Rings
of Saturn

Optical, infrared and radio observations of the A and B rings are well
explained by assuming that particles of centimeter to meter sizes
populate the rings. There are however many indications that smaller
particles are present in Saturn rings, not only in the thin E, F and G
rings but in the thick B ring (spokes) as well. Here, we report on the
presence of subcentimeter icy particles in the A and B rings, based on
the interpretation of far W observations made with the Voyager and IUE
spectrometers. Spectra recorded with these instruments show the strong
ice aborption at ~165U Angstrom. The structure of the spectra cannot be
explained when invoking the absorbing effect of the macro particles long
known to exist in the rings. Instead, smaller particles have to be
injected in the scattering model to produce an adequate W absorption
edge. We will present the size C distributions and associated optical
depths of the various populations of particles that fit the

23.04 Horn L. J.* Cuzzi J. N.
Spatial Structure in Saturn's B Ring

We have analyzed the spatial structure in Saturn's B ring for regional
variations in characteristic spatial scales. We obtained radial scans of
ring reflectivity from the highest resolution full radial scan of the
lit face of the rings in the Voyager 2 data set.

The radial scans were analyzed using a Burg maximum entropy technique to
obtain spatial frequency spectra as a function of radial location in the
B ring. We find that the main rings display characteristic structure on
several scales that vary with location.

The inner B ring (between the inner edge and around 99000 km), has a
characteristic scale of 80-100 km. The B ring optical depth abruptly
increases around 99000 km. Between 99000 and 103500 km, the 80-100 km
scale vanishes and a strong preference for a longer wavelength (200-300
km) is seen. In this region, finer scale structure (20-40 km) is seen in
abundance. This scale structure is absent in the inner B ring. The
region between 103500 km and 116000 km has a very strong preference for
200-300 km scale structure. Spokes are observed in this region and may
have the effect of preventing the development of any fine scale
structure. Outside of 116000 km, fine scale structure again appears in
abundance, on top of the 200-300 km scale, and covers the ring to its
outer edge. The pervasive 100-200 km scale structure could be related to
ballistic transport.

This work was supported by NASA's Planetary Geology and Geophysics

23.05 Ferrari C.* Brahic A.
Arcs in the Encke Gap of Saturn's Rings

A first quantitative study of the azimuthal brightness variations of the
two ringlets embedded in the Encke division of Saturn's rings has been
carried out using Voyagers images. Arcs with sharp edges and clumpy
structure, similar to the Neptune's arcs, are embedded in these
ringlets. Nevertheless they are more widely distributed along the
azimuth and present some hierarchical behalf the azimuthal scale. Few
data on the phase function of the brightness variations provides some
hints on the possible size distribution of the embedded particles. Many
of these arcs, seen by both spacecrafts Voyager 1 and Voyager 2, appear
to be stable over nine months. This gives important constraints on their
orbits and radial widths or on the mechanism that may eventually confine
them azimuthally. These azimuthal structures have a mean orbital motion
compatible with a keplerian motion. Some of them are associated with
local ring distortions as seen in the F ring of Saturn. Waves associated
with the moonlet Pan close encounters may be responsible of the arc
structure into the inner ringlet. The origin of coorbital arcs in the
central ringlet is unclear.

23.06 Showlater M. R.*
Tracking Clumps in Saturn's F Ring

Shown below are complete longitudinal profiles of Saturn's F Ring. The
upper panel was acquired from several outbound Voyager 1 images (phase
angle alpha ~ 124 degrees ) and the lower one from Voyager 2 images
(alpha ~ 95 degrees). The radially-integrated brightness (or "equivalent
width" W) is plotted as a function of longitude in a frame corotating
with the ring material. The profiles reveal that the F Ring changed
considerably during the ~9 months between Voyager encounters.
Nevertheless, each profile contains 2-3 small clumps that are <~1
degrees in length and are many times brighter than the ring's mean W. I
will investigate whether individual clumps are transitory or whether
they persisted between encounters. Because the ring consists of strands
separated by ~100 km radially, clumps in different strands may shift
substantially relative to one another in 9 months. Using additional
images, it will be possible to refine the mean motion of each clump and
thereby assess whether it persisted. Improved mean motions will also
make it possible to identify possible resonant relationships between the
ring clumps and the nearby moons. This work was supported by NASA RTOP

23.07 Thiessenhusen K. U.* Spahn F. Esposito L. W.
Capturing of Moonlets at Co-rotational Resonances in Saturn's

In our previous analysis of the Voyager Photopolarimeter-data, we
detected two features close to the 3:2 co-rotational resonances of
Pandora and Prometheus. At these locations, depletions in the matter
density occur.

We assume that gravitational stirring due to larger bodies captured at
the resonances can be a mechanism to produce such depletions. The
gravitational influence of these large particles or small moonlets heats
the ring matter and enhances the velocity dispersion. This leads to a
density decrease.

In this work, we present the modelation of such a process. Firstly, we
have simulated the process of capturing of a moonlet with an initially
instable orbit at the resonance position. This mechanism is possible for
low-ordered resonances only.

In the second part, we have expanded further sunulations of the gravit-
tional influence of a moonlet to ring regions with high matter density.
We have calculated density profiles as a function of the moonlet mass
and compared them with the experimental profiles.

23.08 Horanyi M.* Juhasz A.
The Dust Torus of Mars

We have revisited the problems related to the orbital dynamics of small
dust particles around Mars., that are continuously produced due to the
ongoing micrometeoroid bombardment of the moons Phobos and Deimos. In
our earlier studies we have examined the effects of electromagnetic
perturbations on micron sized dust grains and also the dynamics of
bigger particles and showed that grains with radius alpha ~ 10
micrometers from Deimos will dominate the dust population of the torus.
In these studies we have added several new effects. Now we properly
account for the orbital motion, obliquity and most importantly the
oblateness of Mars. First we will discuss our new analytic solution to
the orbit averaged perturbation equations that is valid for arbitrary
large eccentricities in the case of uninclined orbits and show the
unexpected role planetary oblateness might play by splitting up the
usual orbital periods. We will also show our computer simulation
results. We have found that the strong coupling between the orbital
inclination i, and the ascending node omega, will result in a
permanently tilted dust torus. This torus is azimuthally symmetric about
a tilted axis. The tilt angle of this axis is about 10 degrees and it
lies in the plane that is spanned by the celestial north and the spin
axis of Mars. The maximum 'edge-on' optical depth of the torus is tau ~
2 X 10^-7, that is comparable to the ethereal E-ring of Saturn and the
'gossamer' ring of Jupiter.

23.09 Hamilton D. P.* Baguhl M. Grun E.
Dynamics of Interstellar Dust

One of the very intriguing results of the Ulysses mission was the
discovery that many dust particles detected by the onboard sensor were
of interstellar rather than interplanetary origin. Both the Ulysses and
Galileo results now indicate that interstellar grains dominate the
micrometeoroid flux outside of 3 AU and probably account for a sizable
fraction of the flux measured inside this distance.

The solar system moves through the local interstellar medium at 26 km/s
toward a point specified by ecliptic latitude 2.5 degrees and ecliptic
longitude 252 degrees. Our initial analyses have assumed that
interstellar dust particles travel from this direction along straight
trajectories. Solar gravity and radiation pressure, however, cause such
particles to move along hyperbolic orbits, thereby deflecting and
focusing them. Over much of the Ulysses orbit, both the focusing factor
and deflection angle remain nearly constant yielding a constant
interstellar flux in agreement with the Ulysses measurements. During the
initial three months of the mission, however, the focusing of
interstellar particles was very strong, possible accounting for the
enhanced flux seem during that time.

The realization that interstellar dust is so prevalent in the outer
solar system has interesting implications for faint dusty rings, many of
which are thought to be maintained by impacts of particles onto small
moonlets. If interstellar particles are the dominant contribution at all
grain sizes, then the impact rate of meteoroids onto moonlets needs to
be reassessed. In addition, the interstellar flux to a planetary
satellite varies as both satellite and planet move along their
respective orbits. The resulting short and long period variabilities
should be observable by Galileo, soon to become the first artificial
satellite of Jupiter.

DPH acknowledges the support of an NSF/NATO postdoctoral fellowship.

23.10-T Baines K. H. Yanamandra-Fisher P. A. Lebofsky L. Momary T.
Near-Infrared Imaging of Saturn's Rings

Near-infrared images of Saturn and its rings were recorded from 1.6
micrometers to 4.9 micrometers near opposition in 1993 and 1994, in
various CH(sub)4 and water absorption bands at NASA/IRTF with ProtoCam
and NSF-CAM . Very little flux emanates from the planetary disk in the
CH(sub)4 bands, hence only the rings are visible; the rings, being
primarily water ice, are strongly absorbing at 3 micrometers and appear
dark, while Saturn's disk remains visible. Photometrically calibrated
images of the rings at 1.75 micrometers, 1.80 micrometers, 2.27
micrometers, 2.33 micrometers, 3.255 micrometers and 3.6 micrometers are
presented. At these wavelengths, Saturn's disk is dark and only the
rings were imaged. Azimuthal brightness asymmetry has been observed at
visual wavelengths; our observations indicate that brightness asymmetry
exists in the infra-red between the east and west ansae of the main
rings. Our preliminary examination of these images indicates that at
3.255 micrometers, the east ansa of the A-ring is about 6% brighter than
the west ansa, but the east ansa of the B-ring is about 9% dimmer than
the west. At 2.27 micrometers, the east ansa of the A-ring is about 6%
dimmer than the west, while both ansae of the B-ring are about the same
brightness. These brightness variations may be indicators of
compositional, and structural differences in the rings. Currently, we
are developing a model that will include the effects of particle shape
and composition in an effort to better understand the brightness
asymmetry between the ansae.

23.11-P French R. G. Roques F. Nicholson P. D. Maene S. A.
Mason E. C. McGhee C. A. Bouchet P. Matthews K. Mosqueira I.
Earth-based Detection of Uranus' Gamma Ring

We observed the 11 July 1992 occultation of U103 by the Uranian rings
with the 5-m Mt. Palomar telescope, the 2.2-m ESO telescope, and the 4-m
telescope at CTIO, using high-speed infrared photometers at Lambda 2.1
micrometers. The Palomar and ESO ring event times were incorporated into
an orbit model for the Uranian rings, allowing the radius scale of the
U103 event to be determined to sub-km accuracy. From CTIO, a sharp dip
in the stellar signal occurred during ingress at a radius of 50026.89
+-0.39 km, compared to the Lambda ring radius of 50023.87 +- 0.21 km
from Voyager occultation data, assuming a circular orbit. The figure
below shows the CTIO chart record of the Lambda event, along with delta
ring ingress (I) and Lambda egress (E) profiles for comparison. The
equivalent width E of the CTIO feature is 0.3 +- 0.1 km well above the
noise level sigma(E) = 0.06 km, and (except for the 'classical' rings)
the only ring-like feature in the entire dataset. The high SNR, the
uniqueness of the event, and its orbital radius all suggest that this is
a detection of the Lambda ring. (From nearby ESO, the SNR was lower,
with sigma(E) = 0.34 km, comparable to the CTIO detection.) No
comparable features were seen from Palomar, or from CTIO during egress,
indicating that the ring is azimuthally clumpy. Supported by NASA Grant

SESSION 24 ....... Chiron Approaching Perihelion
Thursday, 10:30 - 12:00 Crystal Ballroom B
B. J. Buratti and R. L. Marcialis, Moderators

24.01 Stern A.*
2060 Chiron: An Emerging View of an Active World

Object 2060 Chiron was discovered during Kowal's ecliptic survey in
1977. It orbits the Sun in an elliptical, inclined orbit with a period
of 51 years and a perihelion near 8 5 AU. Studies have revealed that
Chiron has a diameter roughly two orders of magnitude larger than
typical comets, displays coma activity on a variety of timescales, and
resides in a short-lived planetary-crossing orbit. I will review the
observational results obtained on Chiron from groundbased and HST work,
speculate on its origin and possible relationship to the Kuiper disk and
point out some important observational and modeling questions that beg
new work as Chiron's upcoming (Spring 1996) perihelion apparition

24.02 Campins H.* Telesco C. Osip D. Rieke G. Rieke M. Schulz B.
The Color Temperature of (2060) Chiron: A Warm and Small

We present three sets of thermal-infrared observations of (2060) Chiron,
obtained in 1991, 1993 from the NASA Infrared Telescope Facility in
Hawaii, and in 1994 from the Multiple Mirror Telescope in Arizona. These
observations allow the first estimates of the color temperature of
Chiron as well as refined estimates of the radius and albedo of its
nucleus. 10/20 micrometers color temperatures of 126+-8 K and 137+-12 K
are obtained from the 1993 and 1994 observations, respectively. These
temperatures are consistent with the Standard Thermal Model (STM,
Lebofsky and Spencer 1989) but significantly higher than those predicted
by the Isothermal Latitude Model. Our estimates of Chiron's radius based
on the STM are in agreement with each other, with the observations of
Lebofsky et al. (1984), and with recent occultation results (Buie et al.
1993). We obtained values for the radius of 74+-12 km (19.2 micrometers)
in 1991, 88+-11 km (10.8 micrometers) and 104+-10 km (19.2 micrometers)
in 1993, and, 94+-7 km (10.6 micrometers) and 91+-13 km (20 micrometers)
in 1994. These values for Chiron's radius, and that based on the 1983
observation of Lebofsky et al. 1984, are plotted in the figure.

24.03 Meech K. J.* Buie M. W. Samarasinha N. Mueller B. E. A.
Belton M. J. S.
HST Observations of Chiron's Inner Coma--A Possible Bound

In order to resolve the question of whether Chiron possesses a bound
dust "atmosphere", we have obtained a series of 44 images of 2060 Chiron
totalling 5,280 sec with the HST Planetary Camera on 1993 February 22,
23 and on March 8, near its minimum geocentric distance (Delta ~ 8.35
AU). We have employed the CLEAN deconvolution techniques to both a
series of simulated Chiron images to test the procedure, and to the HST
data and have found that there is an azimuthal structure in the inner
coma at a projected radial distance of 0.2 arcsec (~1,200 km) from the
nucleus. This also shows up as a change in slope in the radial surface
brightness profile. Near simultaneous ground-based photometry of the
outer coma showed a steep surface brightness profile with a logarithmic
gradient of -2.5, which agreed with the HST data beyond 0.2 arcsec. The
outer coma is reddened with respect to the sun, which was consistent
with scattering from micrometer sized grains. If the change in surface
bnghtne,ss profile slope is interpreted as an exopause boundary, a mass
for Chiron's nucleus is inferred for a given grain size. When combined
with recent nucleus size estimates from occultation measurements (Buie
et al:) and infrared detections (Campins et al, 1994, AJ submitted),
this implies a very low density nucleus near 400 kg m^-3. The Figures
below show a simulated Chiron image and the deconvolved 2/22 composite
image. This work was supported by grants from STScI (GO-3769.01-9lA),
and NASA (NGL 12-001-057).

24.04 Womack M.* Stern S. A.
Searches for CO and HCN in the Coma of 2060 Chiron

We performed millimeter-wave observations of 2060 Chiron to search for
the C0 J = 2-1 and HCN J = 1-0 rotational transitions. The C0
observations were made at 230 GHz during 25-26 February 1994 UT using
the Caltech Submillimeter Observatory (CSO) 10.4-m on Mauna Kea with
spectral resolutions of 50 and 500 kHz. We achieved 3 sigma upper limits
for the C0 2-1 emission of 75 and 35 mk for the 50 and 500 kHz
resolutions, respectively. Assuming (i) Chiron's coma fills the 30" CSO
beam and (ii) the gas excitation, rotational, and kinetic temperatures
are all between 10 and 50 K, then for both resolutions we find
production rate upper limits of Q(CO) < 7-15 x 10^27 s^-l if CO is a
parent molecule, and < 9-36 x 10^27 s^-l if CO is a product of an
H(sub)2CO parent. If the coma is not expanding isotropically, and
instead the CO J = 2-1 line widths are much narrower (e.g., about 1/3
the coma expansion velocity), then the derived column densities and CO
production limits could be up to five times lower. The HCN J = 1-0
observations were made at 89 GHz using the National Radio Astronomy
Observatory (NRAO) 12-m telescope on 8-13 July 1994 UT with spectral
resolutions of 30, 49 and 100 kHz. The spectra will be presented and
constraints on models of Chiron's atmosphere discussed.

24.05 Marcialis R. L.* Hubbard W. B. Hill R. Bus S. J.
Elliot J. L. Olkin C. McDonald S. Foust J. Sopata L.
Bandyopadhyay R. Meserole R. Buie M. W. Spencer J. R.
Wasserman L. H. Millis R. Dunham E. Beichman C. A.
Jarrett T. H. Young J. Ford C. Herter T.
The 1993 Nov 07 Occultation of Ch02 by 2060 Chiron

The 1993 November 07 appulse of 2060 Chiron to the star ChO2 (Bus et al.
1994, A.J. 107, 1814) was observed by five teams with high-speed
photometers (Buie et al. 1993, IAUC 5898). An event lasting 4.5-6.0sec,
seen by the southernmost station (Tierra del Sol, CA) was the only clear
solid-body occultation. A single chord yields only a minimum diameter
(136-181 km), based on the relative sky-plane velocity of Earth and
Chiron. However, assuming the Palomar station observed a grazing event
ISxes the nucleus size at 166 km. This value compares favorably with the
182 +/- 8 km determination based on thermal infrared measurements and
the STM by Campins et al. (1994, submitted to Icarus) in 1991-94, and
168+/- 40km by Lebofsky et al. in 1984 (Icarus 60, 532), as re-analyzed
by Sykes and Walker (1991, Science 251, 777).

A puzzling aspect of the Tierra del Sol observations is that the flux
drop was not total, but only 80%. Proposed explanations for the residual
20% flux are either that the star Ch02 is double, or that differential
refraction of starlight around the limb by an escaping atmosphere
precluded total stellar extinction. The latter hypothesis is preferred
based on subsequent spectral analysis of the star.

Overall attenuation, presumably due to coma, also is seen near the time
of the occultation in at least three of the data sets. These broad,
shallow features decrease with increasing chord distance from Chiron.
Sharp, unresolved dips (<lOkm) iIl the Palomar, Table Mtn., and Mt.
Hamilton light curves possibly are due to a dust jet, but this
explanation is far from conclusive.

24.06 Elliot J. L.* Olkin C. B. Dunham E. W. Ford C. Gilmore D. K.
Rank D. M. Temi P. Kurtz D. Bandyopadhyay R. M. Bus S. J.
McDonald S. W. Barroso J. Lazzaro D. Lopes D. F. Barucci A.
Bosh A. S. Buie M. W. Millis R. L. Wasserman L. H.
Dahn C. C. Foryta D. W. Hubbard W. B. Marcialis R. L.
Reitsema H. Sicardy B. Stone R. P. S.
Material Near the Nucleus of 2060 Chiron from Stellar
Occultation Observations

The star ChO8 (R~11.5) was identified as a possible occultation
candidate for Chiron last year by Bus et al. (AJ 107, 1814). This event
was of particular interest because the star is much brighter than
previous star occulted by Chiron in 1993 (Marcialis, et al., this
conference) which yielded puzzling results. Since the radius of Chiron's
nucleus would subtend an angle of only about 0.01 arcsec, the prediction
for this event was refined with astrometry from CCD strip scans taken at
Lick Observatory. Observations were attempted from the KAO, SAAO,
Observatorio do Pico-dos-Dias, and three portable telescopes deployed in
Brazil. Due to weather problems, light curves of the occultation were
obtained only from SAAO and the KAO (simultaneously with visible and IR
array detectors). The chords probed with the KAO and SAAO were only a
few kilometers apart.

From an interpolation of measurements taken before and after the event
at the USNO, the closest approach of the observed chords to Chiron's
center of light w?s found to be 96 +\- 16 km, but no nuclear occultation
was observed. Occultations by other material were recorded, however. The
sharpest feature has a maximum optical depth of 0.92 +\- 0.02 at visible
wavelengths and a projected width perpendicular to the line of sight of
less than 10 km. Another feature has an average optical depth of 0.12
+\- 0.02 and a width of about 63 km. These appear to be embedded in
material of even lower optical depth that extends for over 600 km
(approximately symmetric with respect to Chiron's center of light). One
interpretation of these results is that the narrow features are due to
jets, while the broader feature is due to coma material. Further
analysis and interpretation of these results will be presented by Olkin
et al. (this conference).

24.07 Olkin C. B.* Elliot J. L. Dunham E. W. Ford C. Gilmore D. K.
Rank D. M. Temi P.
Further Analysis of the Ch08 Occultation by 2060 Chiron Using
the KAO Observations

KAO observations of the ChO8 occultation probed the region near Chiron
with very high spatial resolution (8.9 km in the optical and 17.8 km in
the IR). In the higher signal-to-noise optical light curve (S/N = 52)
three features are evident, while only the deepest feature is seen in
the IR data (S/N = 5.8). This deep feature occurred less than 3 seconds
before the closest approach to the star at 112 + 16 km from Chiron (from
USNO astrometry, see Elliot et al. this conference). In a 1-second
integration, the optical depth of the deep optical feature is 0.52 +/-
0.01, while in the IR it is 0.68 +/- 0.24. Simple models of the features
will be presented. This work supported in part by NASA grant NAG 2-903.

24.08 Dones L.* Levison H. F. Duncan M. J.
Long-Term Integrations of Chiron and Pholus

Chiron and Pholus are the largest known Saturn-crossing objects. Their
dynamics is of interest because they may provide the 'missing link'
between the Kuiper Belt and the short-period comets, which are generally
only discovered within about 2 AU of the Sun (Shoemaker and Wolfe 1982).
Further, as first noted by Hahn and Bailey (1990), some 'Chirons' can
even evolve onto Earth-crossing orbits. We have simultaneously
integrated the orbits of the four giant planets and about 500 test
particles with initial orbits very similar to those of Chiron and
Pholus, using the RMVS code of Levison and Duncan (1994). The 'Chirons'
are followed until they are ejected from the solar system, strike the
Sun or a planet, or survive 50 Myr. Our general findings include the
following: (1) These objects follow highly chaotic orbits which
typically diverge qualitatively in < 10^4 years. (2) For both Chiron and
PhoIus, the most likely ultimate fate is ejection from the solar system
by Jupiter, in a median time of 1.0 Myr (Chiron) or l.5 Myr (Pholus).
Neither body gets closer than 3 AU to Jupiter at the present epoch. (3)
The orbits evolve considerably before ejection occurs; at the time of
ejection, the typical orbit of each has semi-major axis a~200 AU,
perihelion q near 5 AU, and inclination near 10 degrees. The long-term
evolution at large a can be modeled by a random walk at constant q
(Yabushita 1980, Duncan et al. 1987). (4)Many objects spend periods in
which they follow low-eccentricity (0.01 < = e < = 0.1), planet-crossing
orbits; this type of evolution appears to be a promising possible origin
for objects such as Comet Shoemaker-Levy 9. In addition, some objects
evolve onto near-circular orbits roughly midway between two giant
planets. (5) About 20% of the objects spend time in orbits with
perihelia less than 2 AU, including a number of bodies which cross the
orbits of Earth and Venus, for a median time of ~ 10^4 years. (5) The
number of survivors after 50 Myr is orders of magnitude higher than the
extrapolation from 1-Myr integrations would suggest.

24.09-P Dunham E. W. Ford C. H. Stone R. P. S. McDonald S. W.
Olkin C. B. Elliot J. L.
Occultation Predictions Using CCD Strip-Scanning Astrometry

We are developing the method of CCD strip-scanning astrometry for the
purpose of deriving reliable advance predictions for occultations
involving small objects in the outer solar system. We are using a camera
system based on a Ford/Loral 2K x 2K CCD with the Crossley telescope at
Lick Observatory for this work. The columns of the CCD are aligned
East-West, the telescope drive is stopped, and the CCD is clocked at the
same rate that the stars drift across it. In this way we obtain
arbitrary length strip images 20 arcmin wide with 0.58" pixels. Since
planets move mainly in RA, it is possible to obtain images of the planet
and star to be occulted on the same strip well before the occultation

The strip-to-strip precision (i.e. reproducibility) of positions is
limited by atmospheric image motion to about 0.1" rms per strip.
However, for objects that are nearby in R.A., the image motion is highly
correlated and their relative positions are good to 0.02" rms per strip.
We will show that the effects of atmospheric image motion on a given
strip can be removed if a sufficient number of strips of a given area
have been obtained. Thus, it is possible to reach an rms precision of
0.02" per strip, corresponding to about 0.3 of Pluto or Triton's angular

The ultimate accuracy of a prediction based on strip-scanning astrometry
is currently limited by the accuracy of the positions of the stars in
the astrometric network used and by systematic errors most likely due to
the optical system. We will show the results of the prediction of some
recent occultations as examples of the current capabilities and
limitations of this technique.

This work was supported in part by the NASA Airborne Astronomy
and Planetary Astronomy Programs.

Thursday, 1:30 - 2:15 Crystal Ballroom
M. F. A'Hearn, Moderator

Thursday, 2:15 - 3:00 Crystal Ballroom
M. F. A'Hearn, Moderator

SESSION 25 ....... Ida
Thursday, 3:30 - 5:40 Crystal Ballroom A
T. B. McCord and J. M. Sunshine, Moderators

25.01 Belton M. J. S.* Chapman C. Davies M. E. Greenberg R.
Klaasen K. Thomas P. Byrnes D. D'Amario L. Synnott S.
Merline W. Storrs A. Zellner B.
Determination of the Orbit of (243)1 and the Mass of 243 Ida

M.J.S. Belton (NOAO), C.Chapman (PSI), M.E.Davies (RAND), R.Greenberg
(U.Arizona), K.Klaasen (JPL), P.Thomas (Cornell), D. Byrnes, L.
D'Amario, S. Synnott (JPL), W. Merline (PSI), A.Storrs (STScI),
B.Zellner (Georgia Southern U.)

Galileo SSI (Solid State Imaging) data has yielded 46 positions of
(243)1 relative to 243 Ida at 18 time steps over a time interval of 5.4
hours. SPUD shape models and geodesic methods were used to estimate the
body centers of the resolved and irregularly shaped objects. JPL/Galileo
SPICE ephemerides were used to determine the position of the spacecraft
relative to Ida. Observations were also made with the Hubble Space
Telescope (HST) in an attempt to obtain supplementary data on the orbit
of Ida's moon. However, (243)1 was either too faint or too close to Ida
to be seen in Planetary Camera images. Neither the HST or SSI data
showed any objects other than (243)1 in orbit around Ida. Analysis of
the SSI data yields a range of prograde osculating two-body orbits
depending on the assumed mass for Ida. The rms deviations between these
orbits and the observations, over a wide range of masses, indicate that
the mass is not well determined at the accuracy to which the
observations have been processed so far. The reason for this is that
most of the observations are nearly in the orbital plane of (243)1
making the curvature of the orbital path difficult to measure. Further
analysis of the SSI data may mitigate this problem. Well determined
orbital elements include the inclination of (243)1's orbit to Ida's
equatorial plane and the orbital node. Referenced to the spin pole
determined from SSI data (Davies et al. 1994), the inclination to Ida's
equator is 9 deg. Theoretical consideration of the stability of prograde
orbits about Ida restricts the acceptable range of possible orbits.
Details of these will be presented together with our best estimate of
the mass and density of 243 Ida.

Davies et al. (1994) DPS abstract, this meeting.

25.02 Davies M. E.* Colvin T. R. Belton M. J. S. Thomas P. C.
Veverka J. Galileo Imaging Science Team
The North Pole Direction and the Control Network of the
Asteroid 243 Ida

In anticipation of the Galileo flyby, Binzel et al., 1993 published a
physical model of Ida based on analysis of groundbased photometry. They
concluded that Ida was in retrograde rotation with a siderial period of
4.633632 hours and a north pole direction of R.A. 258.75 degrees,Dec. +
30 degrees or R.A. 50.25 degrees, Dec. + 76 degrees (J2000).

When the first Galileo images of Ida were returned, it was obvious that
the asteroid was in retrograde rotation. Also, we knew that it would not
be possible to improve on the value of the rotation period because it
was a very fast flyby.

A preliminary control network of Ida has been computed based on 229
measurements of 64 points on 16 frames. The highest resolution frame
(FDS 0202562778) was taken from a range of 2469 km and had a pixel size
of 25 m. The lowest resolution frame in the network (FDS 0202556000) was
taken from a range of 50415 km and had a pixel size of 504 m. These
frames were separated in time by 1.143 hours, and Ida had rotated 88.8
degrees. The coordinate origin was defined on two frames. In the
computation, the camera twist angle (C3) was not a variable. The
variables were the latitude, longitude, radii of t,he points, the
pointing angles of the camera (C1, C2), and the direction of the
asteroid's axis of rotation. The overdetermination was 2.03 and the
standard error of measurement was 0.006 micrometers (about one-half
pixel). The solution for the north pole direction was R.A. 338.44
degrees +\- 3.5 degrees, Dec. 86.85 degrees +/- 0.4 degrees (J2000).
This value is only 13.5 degrees from Binzel et al., 1993's second
solution. Considering the very irregular shape of Ida, the Earth-based
predictions were very good.

This research was supported by the Galileo project at JPL.

Binzel, et al., Asteroid 243 Ida: Groundbased photometry and a
Pre-Galileo Physical Model, Icarus 105, 310-325 (1993).

25.03 Veverka J.* Thomas P. Lee P. Helfenstein P. Belton M. J. S.
Chapman C. Klaasen K. Johnson T. V. Harch A. Davies M.
Galileo Imaging Team
Ida's Satellite: What Is It Like?

Galileo's flyby of 243 Ida in August 1993 led to the discovery of a
small satellite some 100 km from the asteroid's center. Forty-six images
of the satellite at 18 different observing times have been played back,
including one multicolor sequence in which the satellite is resolved
adequately (~ 105 m/pxl) and two higher resolution single-color views
(39 and 24 m/pxl).

The satellite, mean radius = 0.7 km, is an elongated, but not angular
body with principal diameters of 1.6 x 1.4 x 1.1 km. In the Galileo
views, the longest axis points in the direction of Ida, and its shortest
axis (rotation axis) is perpendicular to the orbital plane. The spin
rate appears to be slow (>~ 1 day?) and may be synchronous. The
satellite shows no conspicuous sharp edges and is much less irregular in
shape than Ida. Limb profiles are remarkably smooth over distances of
200-300 m. The geometric albedos of the two objects are similar (0.22 vs
0.24 +/- 0.03), as are the 0.4-1.0 , micrometers colors. (243)1 has a
slightly deeper 1-,micrometer band than Ida (by 5-8%). While no
identical regions (in color) are seen on Ida, the color difference is
consistent with color variations reported within the Koronis family and
may be due to a slightly higher Pyroxene/Olivine ratio on the satellite.

Almost a dozen craters ranging from <~ 90 to 280 m diameter are visible
in the best image (39 m/pxl at 47 degrees phase). The largest appears to
contain a Positive Relief Feature (~100 m). The image includes an
intriguing crater chain, but no grooves, ridges, or sharp edges are

While the satellite's origin is uncertain, a likely scenario would have
the satellite date from the breakup of the Koronis family. It is
interesting that crater densities on the satellite are similar to those
on Ida itself.

25.04 Thomas P.* Veverka J. Carcich B. Davies M. Belton M.
Chapman C. Galileo Imaging Team
The Unusual Shape of 243 Ida

Galileo images obtained over slightly more than one rotation of Ida on
28 August 1993 cover more than 90% of the asteroid's surface at scales
ranging from 2.4 km/pxl to 25 m/pxl. Control point coordinates, limb,
terminator and shadow positions were used to make a digital shape model
(2 x 2 degrees) of the asteroid. The volume is 16000 km^3 +/- 2400 km^3
(mean radius of 15.6 +/- .8 km). The shape is highly asymmetric,
especially along the long axis. Radii to the center of figure vary by a
factor of 5. A fit ellipsoid has radii of 30.1, 12.6, and 9.3 km. The
spin vector is within 1 degrees of the maximum moment of inertia axis of
the shape model.

Several aspects of the shape are unusual even for small, irregularly
shaped satellites and asteroids: 1) Variations in radii (6-30 km) are
very large; 2) The biggest concavity is not a single facet or bowl but a
25-30 km wide band traversing nearly half of Ida. Included are many
smaller concavities, one of which is about 10 km across and 4-5 km deep;
3) Some sections of the surface are almost radial to the center of
figure. (However, taking into account the direction of local gravity,
this does not mean "vertical" cliffs!) Most importantly, details of
Ida's shape and the character of local topography appear to vary from
region to region, suggesting inhomogeneities in the parent object from
which Ida is derived.

25.05 Sullivan R.* Greeley R. Pappalardo R. Carr M. Kirk R.
McEwen A. Geissler P. Greenberg R. Granahan J. Head J.
Lee P. Thomas P. Veverka J. Moore J. Morrison D.
Galileo Imaging Team
Surface Geology of 243 Ida

Galileo SSI images reveal a variety of morphological detail across the
surface of 243 Ida. Craters occur in a continuum of degradation states
and are generally bowl-shaped, with some exceptions having flat floors,
small central mounds, or straight rim segments. The freshest craters
have narrow raised rims, but distinct ejecta blankets are not visible as
morphologic features. A few of the freshest, brightest craters have one
or two short rays. Some craters have low-albedo floors when viewed under
high sun. Craters >=1 km diameter are five times more abundant on Ida as
on Gaspra, reflecting target strength differences, different surface
ages, and/or different projectile fluxes between these two asteroids.
Lineations are recognized on the basis of albedo (stripes) and
topography (grooves). Stripes are generally brighter than surrounding
terrain and are commonly oriented locally downslope, suggesting
mass-wasting in some cases. Grooves have uniform widths with some
indications of crenulated edges,. and may represent fractures, fissures,
or secondary impact chains. About fifteen isolated positive relief
features have been confidently identified. Isolated positive relief
features are probably boulders originating from impacts, and in some
instances are found within or immediately adjacent to large craters. A
few shallow linear concavities oriented locally downslope within a large
impact crater may represent mass-wasting scars. Color variations across
the surface are generally minor (similar to those reported on Gaspra),
exceptions being small, bright, fresh craters (bluer), some dark floor
craters (redder), and a large diffuse bluer patch of unknown origin.

These observations indicate the presence of substantial regolith on the
surface of Ida. The shape of Ida is considerably more complex than a
triaxial ellipsoid and more irregular than that of 951 Gaspra. This
complex overall shape combined with rugged surface topography imply
considerable variation between local slope and gravity field, suggesting
that regolith should be distributed nonuniformly across the surface.

25.06 Carlson R. W.* Weissman P. R. Segura M. Kamp L. W.
Smythe W. D. Johnson T. V. Matson D. L. Leader F. E.
Mehlman R. Fanale F. P. Granahan J. C. Kieffer H. H.
Soderblom L. A. McCord T. B.
Infrared Imaging Spectroscopy of Asteroid 243 Ida and Discovery
Spectra of Satellite 1993 (243) 1

During the August 28, 1993 flyby of 243 Ida, the Near Infrared Mapping
Spectrometer (NIMS) observed the asteroid using a variety of spectral
imaging modes and geometric conditions. Within the NIMS special range
(0.7 to 5.2 microscopy), the observed radiation arises from both
reflected sunlight and thermal emission. The highest spectral coverage
measurements (408 and 102 wavelengths) show absorption features at 1 and
1.9 um due to iron bearing forms of the silicate minerals olivine and
orthopyroxene. Of these two, the absorption strengths indicate that
olivine is the more abundant species. Measurements obtained at the
highest spatial resolution (1.8 km) obtained near closest approach show
very little spectral variation over the face observed. A spectrum
obtained 2 1/2 hours earlier, corresponding to the average of the
opposite hemisphere, is similar but with larger error limits.

Evidence for an accompanying satellite was discovered in a
102-wavelength NIMS imaging scan. These spatially unresolved spectra,
and other 17-wavelength sub-pixel spectra, show qualitative differences
with those of Ida. Ratio spectra show a relatively narrow absorption
band shortward of 1 micrometers and a weaker, broader band centered at
about 2.3 micrometers. These results suggest that the satellite contains
more pyroxene, relative to Ida, and that it may contain roughly equal
amounts of orthopyroxene and clinopyroxene.

25.07 Weissman P. R.* Carlson R. W. Segura M. E. Smythe W. D.
Matson D. L. Johnson T. V. Leader F. E. Kieffer H. H.
Soderblom L. A. Fanale F. P. Granahan J. C. McCord T. B.
Galileo NIMS Thermal Observations of Asteroid 243 Ida and

The Galileo Near Infrared Mapping Spectrometer (NIMS) observed asteroid
243 Ida during a close encounter on August 28, 1993, at a heliocentric
distance of 2.95 AU. Observations covering one full rotation of the
asteroid were made at a variety of spectral and spatial resolutions. The
best data were obtained in the HiRes observation 3.7 minutes prior to
closest approach at a range of ~3,650 km (in 17 wavelengths between 0.7
and 5.2 micrometers) at a spatial resolution of about 1.8 km per nimsel
(a NIMS pixel), and in the ChemMap observation 7.8 minutes before
closest approach at a range of 6,300 km (in 102 wavelengths between 0.7
and 5.2 micrometers) at a spatial resolution of 3.2 km per nimsel.
Thermal emission from the sunlit hemisphere of Ida was detected at
wavelengths longward of 4.0 micrometers. Analysis of the NIMS data gives
surface brightness temperatures between 180 K and 208 K, assuming an
emissivity of 0.80. The surface temperatures are interpreted as
consistent witli a surface thermal inertia of 0.0015 cal cm^-2 sec ^-l/2
K^-1, indicative of a thick insulating regolith on the asteroid surface.
NIMS also discovered a satellite in close proximity to Ida, detected in
three observations: HiRes, ChemMap, and IdaFin, the latter occurring
16.2 minutes before closest approach at a range of ~12,300 km. Thermal
observations of the satellite, 1993(243)1, will also be reported. This
work was supported by NASA through the Galileo Project at JPL.

25.08 Granahan J. C.* Fanale F. P. Carlson R. W. Kamp L. W.
Klaasen K. P. Belton M. Chapman C. R. McEwen A. S.
Galileo NIMS and SSI Instrument Team
A Galileo Multi-Instrument Spectral View of 243 Ida and 1993

On August 28, 1993 the Galileo spacecraft encountered the asteroid 243
Ida and its satellite 1993 (243) 1 with a variety of remote sensing
instruments. Like Galileo's first encountered asteroid 951 Gaspra, these
objects are S-type asteroids. Unlike Gaspra, 243 Ida and 1993 (243) 1
are members of a well defined asteroid family (the Koronis asteroid
family) and thus provides an excellent comparative data set to study
S-asteroid geology. An asteroid family is a group of asteroids which
possess similar proper orbital elements which may be the result of an
impact disrupted parent body. S-type asteroids are a spectral class of
asteroid which are composed of NiFe, olivine, and pyroxene. The
exact geologic nature of S-asteroids is a controversial topic since this
combination of minerals can be used to describe analogous ordinary
chondrites and many stony iron meteorites. Ordinary chondtites are
primitive solar nebula condensates and stony irons are products of
igneous differentiation.

This study combines the observations of the Solid State Imaging (SSI)
experiment with those of the Near Infrared Mapping Spectrometer (NIMS)
experiment to analyze the mineralogic nature of the 243 Ida system. SSI
gathered a variety of images at wavelengths ranging from 0.4 to 0.99
micrometers and NIMS collected spatially registered spectra at 0.8 to
5.2 micrometers. These SSI/NIMS datasets are also co-registered to other
higher resolution clear filter SSI images to study the correspondence of
spectra differences with motphology. In the highest spatial resolution
SSI/NIMS data set of 243 Ida contains absorptions indicative of olivine
and orthopyroxene plus a "red" spectral slope characteristic of NiFe
with few spectral variations. The few observed spectral variations in
this data set can be attributed to photometric effects and shadows. The
photometric effects are most likely caused by the high angles of
emission of the reflected spectra off of 243 Ida's extreme slopes.
However, there are data of 243 Ida that .were obtained at different
geometries that are still being analyzed that may or may not yield
additional spectral/compositional vatiations on 243 Ida. The SSI/NIMS
data show that 1993 (243) I possesses a S asteroid like spectrum which
contains higher, concentrations of pyroxene than indicated in the 243
Ida data. The variations observed between 243 Ida and 1993 (243) 1 are
still within the spectral variations observed by Binzel et al. (1993,
Icarus 106) amongst the Koronis family members. These cocompositional
measurements further support existing arguments that the Koronis family
parent body had been subjected to igneous differentiation since an
undifferentiated chondtitic parent body is expected to have a
homogeneous mineralogy. Additionally, none of the data returned from the
243 Ida encounter exhibit a "flat" ordinary chondrite; type spectra.

25.09 Veeder G. J.* Matson D. L.
Infrared Comparison of Ida and Other Koronis Family Asteroids

The asteroid 243 Ida has been observed by the Infrared Astronomical
Satellite (IRAS Minor Planet Survey), by the Galileo spacecraft, and at
the Infrared Telescope Facility (IRTF). Ida appears to be a typical
member of the Koronis dynamical family. Ida is well within the limited
range of albedos and infrared colors shown by Koronis asteroids.
Moreover, Koronis family members cluster near the average values for all
S class asteroids. Surface mineralogy (dominated by pyroxene which
implies metamorphism) is independent of size in the present sample. This
uniformity throughout the family reveals no evidence for any
core-mantle-crust structure in the Koronis parent body at this scale.
Thus, it remains ambiguous whether Ida itself was originally part of the
surface or interior. Future work on faint family members may be
comparable to the satellite of Ida. Such higher spatial resolution is
required to address possible small scale length global differentiation.

This work was done at the Jet Propulsion Laboratory, California
Institute of Technology, under contract to NASA. The authors are
visiting scientists at the NASA Infrared Telescope Facility.

25.10 Moore J. M.* Asphaug E. Morrison D. Galileo Imaging Team
Ida Collisional History I: Stress-Wave Focusing as an
Explanation for Grooved Terrain

A ~10-15 km diameter concave facet we interpret to be an impact feature
is observed near Ida's axis of greatest elongation (on the "western
limb"). On the other end of Ida (the antipodal "eastern limb") are a
series of well-resolved linear features whose traces are oriented along
the long axis of the asteroid and resemble the grooves on the martian
satellite Phobos. Recent work based on analytical (Fujiwara 1991) and
numerical (Asphaug and Melosh 1993) models of stress-wave propagation
and fracture have shown how the grooves on Phobos may have grown from
direct and reflected tensile stresses during the formation of the large
crater Stickney. We now examine the possibility that the grooves on Ida
are the result of a similar, though more localized, process. The
"crater" in question is of the same general size as Stickney (~14 km
diameter), but its impact energy was dissipated over a target volume
four or five times as great. Hence, direct impact disruption was not as
severe. But Ida's significant elongation may have caused reflected
tensile stresses to focus acutely along the symmetry axis, leading to a
localized episode of disruption. A 2D numerical model for stress-wave
propagation and fragmentation (Melosh et al. 1992) was used to study the
mode and extent of brittle failure resulting from this impact. From our
model results, we observe that the "eastern limb" grooves antipodal to
the large concave facet on the "western limb" of Ida can be explained in
terms of a homogeneous elongated target impacted near a symmetry axis.
Stresses propagate through the length of the body without causing
disruption, and reflect (as tensile images) to a focus at the antipode.
The greatest region of fragmentation was of course in the vicinity of
the "crater." Damage throughout the rest of the body was minor, except
at the antipode where extensive fracturing also occurred. This process
creates fracture grooves in the regions most distant from the crater
with orientations which match the observations. The target is otherwise
not extensively shattered. Note that a rubble-pile interior prior to the
impact is not consistent with this hypothesis, since seismic wave energy
would be scattered and this focusing phenomenon would not be apparent.
If our model for groove formation is correct, then implicitly the bulk
of Ida must have originated as a single Koronis fragment rather than as
an aggregation of smaller bodies.

25.11 Chapman C. R.* Merline W. J. Davis D. R. Veverka J.
Belton M. J. S. Johnson T. V. Galileo Imaging Team
Ida's Satellite: Its Origin and Impact History

The unexpected existence of Ida's satellite, and its nature as revealed
by high resolution Galileo images, provide constraints on asteroid
collisional history and the impact environment in the asteroid belt. The
little moon (1.6 x 1.4 x 1.1 km, Veverka et al., this meeting) has a
smooth, ellipsoidal shape in striking contrast to the shape of its
parent and of other small bodies that have been imaged. Up to 29 craters
may be visible in the high-resolution image. Diameters of 10 reliable
craters in regions of good visibility (near the terminator) are
consistent with an equilibrium power-law (differential index about -3)
and a surface density consistent with saturation, like the crater
distribution on Ida itself. The satellite's inferred composition from
its reflectance spectrum differs from Ida but is within the range of
Koronis family members (Granahan et al., this meeting).

The satellite's rounded shape suggests that it either (a) has been
eroded (sandblasted) by a steep size-distribution of meters-scale
projectiles or (b) that ithas accumulated into a quasi-equilibrium shape
from small debris in orbit around Ida. Evidence favors the latter.

It is unlikely that Ida's moon was originally an ejecta block from Ida
itself. Probably its precursor was a separate fragment from the
catastrophic disruption of the Koronis family precursor, captured from
its trajectory in a jet (G. Martelli et al., 1993, Astron. Astrophys.
271, 315) that included Ida. The subsequent collisional environment of
both objects (asteroid belt projectiles, perhaps much augmented by
Koronis family fragments compared with the flux to which Gaspra was
exposed) would have been sufficient to disrupt the proto satellite. It
subsequently reaccreted. Its smooth shape and the apparent absence of
other satellites (>300 m diameter) around Ida suggests that it has been
a "sweeper" of ejecta from Ida cratering and of any earlier or temporary
satellites of Ida. A crater chain (3 to possibly 6 craters long) and a
possible block within a prominent crater imply that at least some of the
visible craters on the satellite are due to local projectiles from the
Ida system.

25.12 Geissler P. E.* Petit J.-M. Greenberg R. Bottke W. Nolan M.
Durda D.
Erosion and Regolith Redistribution on 243 Ida and its Moon

Debris generated by impact processes in the Ida system is subject to an
unusual dynamical environment, due to the nonspherical shape and rapid
rotation of the asteroid. We have modeled trajectories of particles
launched from the surfaces of asteroid Ida and its moon in order to
determine the theoretical distribution of ejecta blocks on the surface
of Ida and the eventual fate of possible impact-generated regolith on
the satellite. In our simulations, Ida is approximated as a triaxial
ellipsoid with radii of 28 km x 12 km x 10.5 km, rotating with a period
of 4.6 hours. Ida's moon is considered to be a sphere 1.5 km in diameter
in a stable prograde orbit with small inclination and eccentricity. A
density of 3.5 gram cm^-3 is assumed for both bodies. In each
experiment, large numbers of test particles are launched with a fixed
initial velocity from random surface locations and with randomly chosen
inclination and azimuth directions, and then tracked until they either
reimpact with Ida or its moon, or escape.

Almost all particles launched from Ida at very low velocities (V <<
Vescape) reimpact the asteioid within one rotation period, and their
distribution is random. However, only a few percent of particles
launched at speeds near the escape velocity (~20 ms^-l for Ida)
eventually return to the surface, and their spatial distribution is
markedly nonuniform, with clusters on the asteroid's rotational leading
surfaces (regions which sweep the space before them as Ida rotates).
These particles tend to be launched from rotational trailing surfaces
into retrograde orbits around Ida for periods which are long in
comparison to the asteroid's rotational period. The distribution of
reimpact sites in these models closely resembles the distribution of
locations of large (50-150 m diameter) blocks on Ida's surface, as
mapped by the Galileo Imaging Team [1].

Although few particles launched from Ida collide with its moon, the
reverse is not true. The escape velocity from the satellite is estimated
at only 1 ms^-l, and particles launched at this speed from the satellite
have a substantial probability of colliding with Ida, regardless of
their initial launch direction. Preliminary estimates of the disparity
in regoli? transport suggest that as much as an order of magnitude more
material may be shed onto Ida by its moon than is captured from ?e
asteroid by the satellite.

Reference: 111 Belton et al., First Images of 243 Ida, submitted to
Science 1994.

25.13 Petit J. M.* Greenberg R. Geissler P.
Orbits Around a Small, Highly Elongated Asteroid: Constraints
on Ida's Moon

Discovery of a moon of Ida has raised questions of how it formed and how
long it could survive. If the satellite is a fragment ejected from Ida
after a collision that created a crater, similar to production of
surface boulders (Geissler et al. 1994 LPSC and DPS), the orbit must
have changed immediately to avoid any subsequent close approach. Perhaps
the satellite is a large fragment resulting from the disruption of the
same parent body that produced Ida and the Koronis family. In that case,
the satellite must have survived disruptive collisions, and must have
occupied an orbit stable against escape or impact with Ida for as long
as ~Byr.

To study the stability of orbits of a small body around such a highly
elongated primary, we use the same software that we used for the study
of the boulders: One version models the asteroid as having the same
potential field as a contact binary (two rocky spheres). The other (by
B. Chauvineau) assumes the primary is a triaxial ellipsoid with
principal axes 56, 24, and 21 km, density 3.5 g/cm2 and rotation period
4.6 hr. With s?ch a non-spherical figure of Ida and the satellite at a
distance comparable to the size of Ida, orbits can be quite irregular.
The observed distance of the moon is about 100 km from the center of
Ida. We find that orbits with pericenter >80 km are stable and long
lived. Their orbits are essentially precessing ellipses. Prograde
equatorial orbits with pericenter q <70 km generally collide or escape
within a few days, which places a constraint on the orbit of the
observed moon. Retrograde orbits are generally more stable, unless their
q is less than 50 km.

Special classes of stable periodic orbits may help explain a long-lived
satellite. One class of orbit, synchronous at a longitude 90 degrees
from the long axis, was expected to be trapped in the gravitational
harmonic "topography", but we discovered that such orbits are unstable
for a primary as elongated as Ida (similar to results by D.J. Scheeres,
1994, AAS/AIAA meeting). Also, we have discovered a class of stable,
highly elongated, retrograde orbits that are commensurate with Ida's
rotation, such that pericenter is held away from the long axis,
stabilizing the orbit. Other interesting orbits are found when
non-equatorial trajectories are investigated. Stable orbits may play a
role in scenarios for long-term evolution.

25.14 Durda D. D.*
Numerical Models of the Origin of Asteroidal Moons During
Hirayama Family Formation

The discovery of a small natural satellite of the asteroid 243 Ida has
revived interest in the mode of origin of asteroidal satellites. The
spectral similarity of 1993 (243) 1 to its parent, and the difficulties
inherent in producing a moon by the capture of a passing body, suggest
that 1993 (243) 1 was produced either as 1) a large block ejected from
Ida during a collision or 2) from Koronis parent body debris captured by
Ida during the formation of the Koronis asteroid family. Difficulties
exist in understanding how to prevent large collisional debris from
reimpacting the surface of the primary in the first scenario. To test
the second scenario, I will present the results of numerical simulations
of the initial stages of the expanding debris field resulting from the
fragmentation of a parent asteroid. Three dimensional N-body
integrations of several hundred fragments with a power-law distribution
of masses indicate that the formation of co-orbiting fragments and even
contact binaries is indeed possible. The fragments are launched from
their initial locations within the parent body in quasi-radial
directions with a power-law distribution of ejection speeds dependent
upon fragment mass. Results indicating the dependence upon model
parameters as well as the statistics of numbers of orbiting pairs and
possible contact binaries wi11 be presented.

25.15-P Helfenstein P. Veverka J. Thomas P. Johnson T. V. McEwen A.
Granahan J. Fanale F. Geissler P. Belton M. Chapman C.
Galileo Imaging Team
Regolith and Composition of Ida: Clues from Color and

Galileo coverage between phase angles 19 degrees and 95 degrees,
complemented with telescopic data at lower phase angles, indicates that
Ida has a steeper phase curve, particularly near opposition, than does
Gaspra, consistent with a more intricately textured surface (more
complete regolith cover?) on Ida. Although the average geometric albedos
are comparable, 0.24 (Ida) and 0.23 (Gaspra), at phase angles beyond 10
degrees, Ida is generally 25% darker than Gaspra. As on Gaspra, albedo
markings are subdued.

Globally averaged Galileo SSI color data between 0.4 and 1.1 ,
micrometers agree well with telescopic measurements. Small but
significant color variations are observed over the surface, involving
both the blue part of the spectrum (0.40 to 0.56 and 0.76 micrometer
ratios) and the depth of the 1-micrometer olivine/pyroxene band
(0.76/0.99 micrometer ratio). As on Gaspra, color variations on Ida are
correlated with albedo: brighter areas have deeper 1-micrometer
absorptions and "bluer" visible colors. Such a correlation suggests that
the brighter areas may have a slightly higher pyroxene content and/or a
lower metal content, and is generally consistent with a slight degree of
optical alteration in a regolith which is "less mature" on steep slopes
and in fresher craters. The spectrum of Ida's satellite, which is very
similar to but significantly different from Ida's, does not conform to
the pattern above. The satellite's albedo is slightly less than that of
Ida (0.22 compared to 0.24) but its 1-,um band is deeper than those of
even the bluest areas on Ida. The most likely explanation is that the
satellite has a slightly different composition (more pyroxene?) than any
area of Ida. It is also probable that the color differences on Ida
itself are at least in part due to slight differences in composition.
The range of spectral variation seen on Ida and its satellite is
consistent with that reported for members of the Koronis family. It is
unlikely that color variations among similar-sized Koronis family
members are due solely to differences in regolith processes, suggesting
that the Koronis parent body was a differentiated object.

SESSION 26 ....... Outer Planet Satellites
Thursday, 3:30 - 5:30 Crystal Ballroom B
J. R. Spencer and D. P. Simonelli, Moderators

26.01 Coombs C. R.* McCord T. B. Sunshine J.
Compositional Unit Interpretation on the Icy Galilean

Surface units, as defined by multispectral maps from Voyager, are an
important indicator of past processes and the original state of the icy
Galilean Satellites. Spectrometry, as practiced from groundbased
telescopes, has supplied some material identification on a global scale.
Based on these data, distinct spectral units have been found on all the
icy satellites. In the visible there are high- and low-albedo units. The
bright material is composed of ice of differing grain size and is
believed to be endogenic in origin. The dark material differs in origin
and composition. We believe, from our analysis, that the spectral
evidence suggests the presence of at least three different dark
materials, each associated with a different spectral signature: (1) a
hemispherical UV-dark material, (2) a tectonically,related dark
material, and (3) an impact-related dark material. The UV dark material
is distributed in a sinusoidal hemispheric pattem and is most
distinguishable on Europa. Suggestions for its origin, by others and
from our work, include magnetospheric sputtering and ion implantation.
The second, lineament-related, dark material unit appears to be
indigenous to the satellites and may have been redistributed by internal
metamorphic processes. The third material, associated with the impact
craters, may have been derived from outside the Jovian system. Our
current analyses of the Voyager multispectral data and groundbased
spectro-photometric data is helping to further constrain these units,
leading to a better basic scientific understanding of their composition
and spatial distribution, and to help plan the upcoming Galileo Jupiter
orbit measurement phase and to lay a foundation for interpreting the
NIMS and SSI data.

26.02 Domingue D. L.* Lane A. L. Lankton M.
Have the UV Reflectance Spectra of the Galilean Satellites Io
and Europa Changed with Time?

Between May 14-18 we performed near-UV observations of the Galilean
satellites, focusing on Io and Europa, at the University of Hawaii 2.24m
telescope on Mauna Kea. We obtained spectra from 295 to 347 nm at about
0.024 nm resolution using the wide slit (~3 arcsec) of the WSI
spectrograph. The slit captured the entire disk of the satellite with
margin and the leading and trailing hemisphere observations were made at
similar airmasses. Preliminary analysis of our Io and Europa spectra
show interesting results compared to the earlier work of Nelson et al.
(1987, Icarus, 72, 358-380) who observed these satellites using IUE
(1978-1986). Comparing our leading vs. trailing hemisphere ratios of Io
with those of Nelson et al. (1987) we see no spectral slope differences,
however there is a substantial difference in the amplitude of the ratio.
This difference in amplitude between the two ratios indicates either a
relative brightening of the leading hemisphere or a relative darkening
of the trailing hemisphere. There have been no reports of major hot spot
activity during the 6 to 9 months preceding our observations. Assuming
volcanism is the major mechanism for generating S02, then possible
scenarios, which would explain the changes in the spectral ratios
coupled with the decrease in volcanic activity include: a) radiation
darkening of the trailing hemisphere, and b) sublimation from
midlatitudes where the S02 frost (UV dark) is thinner on the leading
hemisphere (revealing brighter sulfur deposits beneath).

The comparisons of our trailing vs. leading hemisphere spectra of Europa
with those of Nelson et al. (1987) show spectral changes in addition to
amplitude changes, as seen on Io.

Work funded in part by the Planetary Astronomy Program, NASA.

26.03 Lane A. L.* Domingue D. L. Price S. K. Lankton M.
A Re-examination of IUE Near-UV Data: Have Ganymede and
Callisto Changed with Time?

The original work from IUE data taken between 1978 and 1985 reported by
Nelson and Lane (1) and Nelson et al. (2) used the data processing
system developed by the IUE Project (1978-1991). During the past 6 years
the processing system has been markedly refined and enhanced, permitting
a re-examination of some data at increased signal-to-noise. Selected
Ganymede and Callisto spectra were reprocessed from the observing period
1984-86, with increased accuracy in wavelength registration and
significant noise reduction. Ganymede's spectral characteristics from
260 to 330 nm show no significant difference from what was published
earlier. Ganymede shows a gradual darkening of the trailing side with
decreasing wavelength (~15%). Nelson et al (1,2) showed Callisto
possessed an almost constant leading-trailing side ratio (within the
noise bounds). The new processing system markedly lowers the noise
envelope and shows several spectral features between 250 and 320 nm, as
well as a possible broad absorption feature (240-330 nm) on the trailing
side. A small sample of near-W data acquired by us at the 2.24m U. of
Hawaii telescope at MKO in May 1994 shows Callisto has a discernible
spectral change over 20 degrees of rotation (undetected in the IUE data)
and that our Ganyrnede data appear similar to the IUE measurements made
in 1978-85.
References: (l)Planetary Satellites, in Exploring the Universe with the
IUE Satellite, 67-99, 1987, Kluwer Academic Publishers; (2) Icarus, 72,
358-380 (1987)

26.04 Calvin W. M.* Spencer J. R.
Identification of O2 on Ganymede

Two new, narrow absorption features were identified on the trailing
hemisphere of Ganymede by Spencer and Calvin (BAAS, 25, 1119, 1993).
These absorptions, at 0.5773 and 0.6275 micrometers are attributed to
the 1-0 and 0-0 double electronic transitions from the ^l delta (sub)g
<-->^3 Sigma (sub)g states of O(sub)2 (i.e., simultaneous transitions in
two interacting O(sub)2 molecules). While observed in the gas at high
pressures, these forbidden transitions have greatly enhanced intensities
in condensed O(sub)2 due to perturbations by intermolecular forces and
are responsible for the blue color of both the liquid and solid (e.g,
Tsai and Robinson, J. Chem. Phys., 51, 3559, 1969). The 1-0 and 0-0
transitions are the strongest (by an order of magnitude) visible
absorptions observed in thin-film transmission measurements of the three
crystallographic forms of solid O(sub)2 (Landau et al., Spectrochim.
Acta, 18, 1, 1962), and an average of these three corresponds extremely
precisely with the absorption features in the spectrum of Ganymede as
seen in Fig. 1.

Given the temperature and surface pressure of Ganymede it is unlikely
that the O(sub)2 is present as a pure solid or liquid. (O(sub)2 requires
a temperature of 90K to liquify and 54K to solidify.) Reguired pressures
for the gas are also unreasonable. Rather, we suggest that O(sub)2 is
trapped either at defect sites in the water ice surface or possibly in
an ice clathrate. The concentration of the absorption features on the
trailing hemisphere is consistent with production through magnetospheric
bombardment and ice irradiation experiments (e.g. Johnson, Ch. 3,
Energetic Charged-Particle Interactions with Atmospheres and Surfaces.

Fig. 1., which appears here in the hard copy shows observed spectrum of
the Ganymede trailing hemisphere ratioed to Callisto from 20/mar/93. The
dotted lines are the average of the absorption from the three different
forms of solid O(sub)2. This average has been scaled to the approximate
observed band depth on Ganymede.

26.05 Showman A.* Stevenson D. J.
Coupled Orbital and Thermal History of Ganymede

Malhotra (1991) proposed that Io, Europa, and Ganymede passed through an
early orbital resonance which pumped Ganymede's eccentricity to
0.01-0.03 and may have caused enough tidal heating to allow resurfacing.
Showman and Stevenson (1994) heuristically suggested a means by which
resurfacing could occur. At the onset of resonance, Ganymede may have
been cold and frozen and thus have had a high Q/k. As the resonance
proceeded and Ganymede's eccentricity rose, dissipation would increase.
This increased dissipation would cause a gradual temperature increase,
and might eventually lead to melting. The decrease in Q/k caused by
these events would cause a runaway, leading to a rapid, massive melting

We foresee two possible resurfacing mechanisms. First, when the
eccentricity is high, tidal heating of local near-surface thermal
inhomogeneities may cause resurfacing, even if the bulk satellite
undergoes very little melting. Second, resurfacing may occur during the
runaway melting episode. The runaway melting would produce a massive
global ocean underlying a thin lithosphere. Cracking caused by the
stress of satellite expansion (a result of melting) may then allow
passage of liquid water or soft ice to the surface. We here explore the
second mechanism by testing the likelihood that runaway melting episodes
of short duration actually occur. We have tested the plausibility of our
heuristic mechanism by coupling the orbital model of Malhotra (1991) to
a depth-dependent model of Ganymede's thermal structure, which provides
a time-dependent Q/k to the orbital model. The thermal model takes into
account the different ice and water layers, and treats heat loss using a
parameterized convection scheme.

We specifically address the following questions: (1) Does Ganymede have
high enough Q/k at onset of resonance to allow significant eccentricity
pumping? (2) Does a runaway melting episode actually occur? If so, what
is the timescale? What magnitude of melting occurs? What is the final
thickness of the lithosphere? (3) Crater counts indicate that the
resurfacing occurred over several hundred million years. However,
preliminary models suggest that the runaway melting episode occurs over
much shorter timescales (as it must to produce massive melting and
cracking). Thus, resurfacing by the runaway melting mechanism may not be
consistent with the inferred resurfacing timescale, if only one runaway
melting episode occurs. Possibly, after a first runaway melting episode,
the satellite again undergoes eccentricity pumping, leading to
additional runaway melting episodes hundreds of millions of years after
the first episode. We address the question: Do multiple runaway melting
episodes occur? The current work does not address the question of how
material actually reaches the surface, but is a necessary first step in
evaluating the hypothesis of resurfacing by resonance. R. Malhotra
(1991) Icarus 94, 399-412. A. Showman and D.Stevenson (1994) abstract,
Icy Galilean Satellite Conference, San Juan Institute, Feb 1994.

26.06 Leliwa-Kopystynski J.* Kossacki X. J.
Evolution of Small Icy Satellites--The Role of Ammonia

This work concern the evolution of the internal structure of the icy
satellites from the beginning of their formation until the present
epoch. The icy satellites are considered here as the bodies which have
been formed by the process of homogeneous accretion from protosatellite
icy and mineral dusty medium. Therefore, the consecutive layers of
icy/mineral mixture deposed on the growing satellite present themselves
the layers with constant icy to mineral mass ratio and with the porosity
depending on the radial distance and decreasing with time due to
rheological processes. Modelling of the satellite internal structure is
based on the time-dependent equations of structure of the spherical
symmetric planet: eqs. of mass, of momentum, and of energy (Kossacki and
Leliwa-Kopystyneki, Plan. Space Sci., v. 41, pp. 728-741, 1993). The
radiative equilibrium with solar radiation consists a boundary condition
on a satellite's free surface; the conductive heat transfer is
considered within a satellite. This set of equations is supplemental
by: (1) standard equation of energy production from decay of radioactive
nuclei, and by (2) the empirical rheological equation describing the
rate of decrease of porosity of the icy/mineral mixture which is
equivalent to compaction rate if the components of mixture can be
considered as non-compressible ones. According to laboratory data
compaction rate decreases by about two orders of magnitude when mineral
powder is admixed to that of water ice in the mass ratio 1:1
(Leliwa-Kopystynski and Meeno, J. Glaciol., v. 39, pp. 843-856, 1993).
On the other hand, compaction rate dramatically increases if the pure
water ice component is replaced by water/ammonia ice (Leliwa-Kopystynski
and Kossacki, Plan. Space Sci., submitted). Calculation of internal
structure of Minas, Miranda, and Enceladus have been done; the mineral
mass to total mass fraction was taken as equal to 0.445, 0.631, and
0.445, respectively; ammonia to water mass fraction was a parameter
varying within an interval from 0 to 0.15. It was found that
two-layered structures develop within the satellites; a low-porous
central area and an outer area where porosity increase to its surface
value. In a case of higher concentration of ammonia the nonporous core
forms. It is possible to select the plausible set of pairs of
parameters (ammonia concentration and surface porosity) by means of
fitting the calculated satellites models to the known satellites size
and moment of inertia. For example, it was found the ammonia
concentration belonging to the range from 0.1 and 0.15 and the surface
porosity equal to 0.5 fit reasonably to the present satellites sizes and
moments of inertia.

26.07 Gordon M. K.* Murray C. D.
Possible Detection of Additional Small Satellites of Saturn in
Voyager Images

There is substantial evidence for the existence of small satellites in
the Saturnian system in addition to those with well determined orbits
derived from ground based and Voyager observations. Synnott (Icarus 67,
189-204, 1986) gives several possible, single detections in Voyager
images and Chenette and Stone (J. Geophys. Res 88, 8755-8764, 1983)
review the evidence from charged particle data for the existence of
material co-orbital with Mimas. We have extended the work of Synnott
using a combination of the NAIF/SPICE Toolkit, IRAF, and locally
developed software to analyze images in the Voyager archive. Possible
new satellites were detected in the images after the identification and
elimination of features such as blemishes and background stars. We give
details of the search criteria and discuss probable orbits for several
candidate objects with diameters ~10-20 km. We note that the 1995 Saturn
ring plane crossing will provide an excellent opportunity to extend the
search with ground based telescopes.

This work was partly funded by the UK Particle Physics and Astronomical
Research Council.

26.08 Petmecky S.* Klavetter J. J.
Near-Infrared Spectrophotometry and Broadband Photometry of the
Saturnian Satellites

Broadband and narrowband near-infrared CCD images of the Saturnian
satellites were obtained using the ProtoCam at the IRTF observatory over
four nights, from August 8 to August 11,1990. The broadband images
included the J, H, K and L filters. The narrowband images were taken at
wavelengths ranging from 2.0 micrometers to 4.3 micrometers at 0.1
micrometer resolution. The 3.4 micrometer feature observed in other
small solar system objects was specifically sought as a possible feature
in the satellite spectra. Image reduction was performed utilizing IRAF,
accounting for bias, and flatfielding. The solar spectral features were
compared with the standard star HDl 160 and solar analog 16CygB.

The analysis of the photometry of the broadband features and the
spectrophotometry of the narrowband features will be presented in this
report. Specific spectral features will be reported and discussed.

26.09 Wilson P. D.* Sagan C.
Spectrophotometry and Organic Matter on Iapetus: Composition

Iapetus shows a greater hemispheric albedo asymmetry than any other body
in the solar system. Hapke scattering theory and optical constants
measured in this laboratory are used to identify possible compositions
for the dark material on the leading hemisphere of Iapetus. The
materials considered are poly-HCN, kerogen, Murchison organic residue,
Titan tholin, ice tholin, and water ice. Three-component mixtures of
these materials are modeled in intraparticle, particle, and areal
mixtures. In a computer grid search of ~2 x 10^7 models, an
intraparticle mixture of 25% poly-HCN, 10% Murchison organic residue,
65% water ice is found to best fit the spectrum, albedo, and phase
behavior of the dark material. The Murchison residue and/or water ice
can be replaced by kerogen and ice tholin, respectively, and still
produce very good fits. Areal and particle mixtures of poly-HCN, Titan
tholin, and either ice tholin or Murchison residue are also possible
models. Poly-HCN is a necessary component in almost all of the good
models. To test the hypothesis that the dark material on the leading
hemisphere of Iapetus comes from Phoebe, the best composition models of
the dark material are used to also model photometry of Phoebe. Although
inconclusive, the best fits to the Phoebe observations have significant
differences from the best Iapetus fits.

26.10 Burns J. A.* Hamilton D. P. Mignard F.
The Contamination of Iapetus by Phoebe Dust

Two decades ago Steven Soter proposed that the very dark leading
hemisphere of Iapetus was coated by material from the low-albedo
satellite Phoebe, whose orbit is retrograde and exterior to Iapetus'.
According to this model, interplanetary micrometeoroids striking Phoebe
release large quantities of micron-sized ejecta, some of which
eventually find their way inward to Iapetus.

We re-examine this hypothesis from our improved understanding of the
effects of solar tidal perturbations and radiation forces on the orbits
of distant circumplanetary dust. The eccentricities of micron-sized
grains launched on circular orbits from Phoebe rapidly increase to
values sufficient for Iapetus-crossing orbits. We demonstrate this
growth analytically for small e and also show it numerically. Although
larger particles launched from Phoebe do not attain these
eccentricities, they ultimately drift past Iapetus due to
Poynting-Robertson drag. We discuss the relative importance of the two
mechanisms, and determine characteristic impact speeds, directions and
timescales. We show how Phoebe particles can reach around the equator
(>180 degrees of longitude), but only partially cover the polar regions.
We note that chaotically tumbling Hyperion and cloud-shrouded Titan will
not exhibit any orbital brightness variations due to this mechanism. We
also suggest reasons why the spectral properties of Iapetus' dark side
might differ somewhat from those of Phoebe.

26.11 Black G. J.* Nicholson P. D. Thomas P. C.
Rotational State of Hyperion

We have numerically integrated the full three dimensional rotation of
Hyperion. The moments of inertia, initial pole position, and spin rate
are from a new shape model and rotation pole for Hyperion which are
based on fitting control points, limb, and terminator positions in high
resolution Voyager 2 images. These images were taken over a 38 hour
period and cover ~ 114 degrees of rotation. From this solution, it is
found that at the time of these images the spin axis was tilted ~60
degrees from the orbit normal and roughly aligned along the axis of
minimum moment of inertia. In addition, the instantaneous spin rate is
found to have been 72 degrees +/- 5 degrees per day, or slightly over 4
times the synchronous rate. The integrated dynamical model using this
solution provides an excellent fit to the lightcurve obtained from low
resolution Voyager 2 images, whereas a fit assuming a constant rotation
pole does not. Previous work by Wisdom, Peale, and Mignard (1984)
predicted that is was likely that Hyperion's large moment differences
and large orbital eccentricity would keep it in a chaotically tumbling
state, and observations by Klavetter (1989) could not be explained by
any simply periodic rotation. However, in these studies spin rates as
high as those found in this work were not investigated. Our integrations
show that this unusual state is quasi-regular, meaning that Hyperion
remains in this state for at least a century, undergoing expected forced
precession, with a period ~/- 300 days, and a free precession (wobble),
with a period of ~/- 15 days. The state is formally chaotic, with the
shortest Lyapunov time on the order of the orbital period or less, but
this chaos appears to lie in the phasing of the spin rate. In this
orientation, Hyperion is found not to tumble on this timescale unless
the spin rate is lowered below ~ 3.6 times synchronous, well below the
uncertainty in the spin rate. Thus we expect Hyperion to still be in
this state today, although the exact rotation and precessional phases
cannot be predicted.

26.12 Tryka K. A.* Brown R. H.
Volatile Transport on the Uranian Satellites

Observations of the Uranian satellites have been able to identify only
water ice on the surfaces of these bodies. Considering that theoretical
calculations have shown that ammonia (Lebofsky, Icarus 25, 205-217,
1974) is stable at a heliocentric distance comparable to Uranus's, it
seems unusual that ammonia, or less volatile substances, such as carbon
dioxide, have not been detected on the surfaces of these satellites.

The high obliquity of the Uranian satellites causes the polar regions to
go through very severe seasonal cycles, and also removes the possibility
that there are permanently shaded areas near the poles to act as cold
traps for volatiles. Because of the geometry of the Uranus system, we
have only been able to study the southern hemispheres of the satellites
in detail, and only during the summer season. This raises the question:
Can volatiles exist on the unobserved winter hemisphere? For the answer
to be yes, there must be seasonal transport of volatiles away from the
summer hemisphere.

We are investigating seasonal volatile transport models for the Uranian
satellites and will report on our progress.

26.13-P Polk K. Hawke B. R. Granahan J. C. Fanale F. P.
Cloutis E. A.
Albedo Anomalies on Ganymede: Implications for Surface
Compositions and Processes

As part of a continuing effort, we have been investigating albedo
anomalies on the surface of Ganymede. Numerous unanswered questions
exist concerning the near-surface stratigraphy of Ganymede as well as
the processes that have produced the various surface units. Previous
studies have demonstrated that important information concerning target
stratigraphy, cratering processes, and projectile populations can be
obtained from investigations in crater morphology and albedo. Hence, we
utilized the Voyager imagery of Ganymede to identify, classify and map
the distribution of unusual albedo variations associated with fresh
impact craters. In addition, a Voyager image/spectral data cube was
constructed and used to better understand the compositions and modes of
formation of these anomalies. An image/spectral data cube is a three
dimensional array in which the x and y coordinates represent spatial
inforrnation and the z coordinate contains spectral data. The Ganymede
data cube consists of the five Voyager band data at a scale of 0.5
degrees/pixel or less. Two versions of the spectral data cube have been
prepared. One is calibrated using Voyager radiance values, and the
second corrects the Voyager values based on ground based observations.

The classes of albedo anomalies include: 1) asymmetric bright ejecta
deposits, 2) dark-haloed craters, 3) dark-rayed craters, and 4) bright
rays or ejecta deposits which,appear to stop abruptly at geologic
boundaries. Numerous examples were found where bright rays which
originated at craters located on grooved terrain disappeared at terrain
boundaries. Our results indicate that these rays become indistinct on
the dark, cratered terrain because the ice-rich primary ejecta was
contaminated by low albedo, near-surface material during secondary
cratering events. A similar process could account for the abrupt
termination of bright ejecta deposits at grooved terrain/cratered
terrain contacts and for asymmetric ejecta deposits associated with
craters which straddle these contacts. Other asymmetric ejecta deposits
were produced by oblique impact. The dark deposits surrounding some
impact craters appear to have been produced by the excavation of buried
low albedo material. In other instances, projectile contamination
appears to be responsible for the production of dark exterior deposits.

26.14-P Consolmagno G. J.
Effect of an Insulating Surface on Ganymede's Thermal History

A number of finite-difference time dependent computer models for
Ganymede's thermal and physical evolution have been run, varying the
thermal properties of the upper 7-30 km of Ganymede. They demonstrate
that given reasonable values for surface conductivities, even simple
models can reproduce the major features of Ganymede's thermal history
implied by studies of its surface geology.

In these models it is assumed that Ganymede started out melted and
differentiated. The temperature throughout both the liquid mantle and
rocky core is set at 300 K while the surface temperature is fixed at 100
K. Heat is generated by the decay of chondritically abundant long-lived
radionuclides in the rocky core, and flows through the core and water
mantle to the surface (where the temperature is fixed at 100 K). The
formation of high pressure phases of ice is calculated, with appropriate
conductivities and latent heats, and convection in the rocky core and
both liquid and frozen ice phases is modelled by a parameterized
convection scheme.

If no insulating layer exists, Ganymede would be completely frozen
within 800 million years, and vigorous near surface convection would
cease after 1 billion years. On the other hand, if the average thermal
conductivity of the upper 7 km is half that of ice, or if there is a
deeper regolith (30 km thick) with roughly 75% the conductivity of solid
ice, then molten regions within Ganymede could last over one billion
years and vigorous near-surface convection could persist for two billion

By inspecting grooved terrain, crater ejecta, and the deformation of
craters of different sizes, various workers have concluded that
Ganymede's upper surface is broken up or inhomogeneous to a depth of
5-25 km. The lithosphere was relatively thin and deformable, and
occasional resurfacing events possible, up to about 2 billion years ago;
then in a relatively short period of time the heat flow dropped
dramatically. These features can be reproduced even in this simple
thermal model.

26.15-P Wagner R. J. Neukum G. P.
Surface Units on Callisto: A Pre-Galileo View

The densely cratered surface of Callisto provides a clue to geological
conditions in the early jovian system. Based on map-projected and
contrast enhanced images, photogeologic m in a scale of 1: 10,000,000
from the Valfoedr and Igaluk regions of Callisto have been compiled. The
latter region will be a primary target of high resolution imaging during
the Galileo flybys. At least three units were mapped: heavily cratered
material, dark cratered material, and smooth material. Smooth, less
densely cratered material indicates locally restricted zones of
resurfacing. Three types of craters may be distinguished: fresh crater
material, degraded crater material and heavily degraded crater material.
Crater size-frequency measurements have been carried out on the units.
As crater density has not reached saturation level at least down to a
diameter of about 8 km, crater retention ages of all units may be
determined. Independence of crater retention ages from the distance to
the apex point of orbital motion as well as crater scaling relationships
indicate velocity impacts derived from primarily planetocentric orbits
for the major part of the projectiles. Additional analyses of lineament
systems show that the Callistoan surface has not undergone any major
surface modification since the time of early heavy bombardment.

26.16-P Le Louarn M. Glenar D. A. Hillman J. J. Drummond J. D.
Disk-resolved Photometry of the Galilean Satellites and Titan
Using AdaptiveOptics

The Galilean satellites and Titan were imaged at wavelengths from 640 to
970 nm, over a range of orbital phase. We observed J1-J4 with the goal
of detecting spatial variations in the surface albedo which contribute
to previously measured leading - trailing side reflectivity ratios, and
also as a check for consistency with recent HST images by Sartoretti et
al. (to be published in ICARUS, '94). Titan was imaged in the 940 nm
atmospheric window with the goal of detecting albedo variations.
Observations were made using the Kirtland Air Force Base, 1.5 meter
adaptive optics telescope near Albuquerque, NM. Using a combination of
tip-tilt and high-order laser guide star correction, it is possible to
achieve spatial resolution on these objects which approaches the
diffraction limit. The camera is a tunable, acousto-optic.imaging
spectrometer developed at GSFC with a spectral resolution of about 2 nm.
Spatial sampling at the 1.5 m Coude feed was about 0.12 arcsec per
pixel. The images were photometrically corrected and reduced to relative
reflectivities using calibration stars and standard reflectance areas on
the Moon. We summarize the results of these observations.

This work was supported by the NASA Planetary Instrument Definition and
Development Program and the GSFC Director's Discretionary Fund.

26.17-P Matheson P. L. Shemansky D. E. Noll K. Weaver H.
Observational Limits for OH in the Magnetosphere of Jupiter as
Inferred from HST FOS Observations

Several planetary satellites in the solar system are covered primarily
by thick layers of water ice. Many of these, particularly the icy moons
of Saturn and Jupiter, orbit within a region in which their surfaces are
bombarded by large fluxes of energetic particles that can loose a steady
supply of neutral gas into the magnetosphere. With the discovery of
large amounts of OH near Tethys, it has become increasingly clear that
gas produced in this way can have a significant effect on magnetospheric
plasma processes. We report here on the effort to observe OH in the
magnetosphere of Jupiter with the Hubble Space Telescope Faint Object
Spectrograph (FOS) as sourced from Europa and discuss the likely role of
OH in the dynamics of the jovian magnetosphere. Observations in the
orbital plane of Europa are scheduled for late August, 1994, while Noll
and Weaver have observations near the limb of Europa scheduled for early
July and August, 1994.

Water-product neutrals in a magnetosphere affect the plasma in a variety
of ways. Indeed, at Saturn, the water-based neutral gases from the icy
moons dominate the plasma as the primary source of magnetospheric mass
and as the primary electron energy loss process throughout radiative
losses (Shemansky, et al., Nature 363, 341, 1993). H(sub)2O, and OH
quickly dissociate under low-energy electron bombardment into their
atomic constituets. Atomic oxygen is particularly effective in depleting
protons from magnetosheric plasma through charge exchange. Europa has
long been thought to provide a neutral cloud of OH into the volume of
space around it, but the evidence for its existence has been lacking,
due to unavailability of space-based instruments, and in part due to the
overwhelming dynamics of the nearby Io plasma torus. Only recently has
reanalysis and revamping of magnetospheric models based upon Voyager PLS
and UVS data given substance to the claim of a Europa plasma torus
(Bagenal, et al., Geophys. Res. Let., 19, 79-82, 1992). Calculations
currellt as of this abstract's submission indicate that an OH neutral
cloud in the orbital plane of Europa could be as bright as 5 Rayleighs,
with OH densities has high as 3 cm^-3.

26.18-P Shi M. Westley M. S. Baragiola R. A. Johnson R. E.
Magnetospheric Plasma and UV Photon Sputtering and Modification
of the Surfaces of the Icy Satellites

Recently we reported the first absolute measurements of the sputtering
of low temperature water ice by Lyman-alpha photons [1], and larger than
expected yields for the sputtering of water ice by keV O^+ [2]. Here we
report new measurements of the sputtering of water ice by UV photons and
keV ions relevant to planetary magnetospheres in the outer solar.
Because we found that the photo-sputtering yields exhibited pronounced
irradiation dose and temperature dependences, we have examined the
synergism between plasma and photo-sputtering of the icy satellites in
the outer solar system. (Raul even if the synergism experiments are not
completed, we can probably roughly estimate the size of the effect.) We
have used these results to reevalute the plasma sources in Saturn's
inner magnetosphere and can account for significant increases over our
lower limit estimates presented earlier [3]. We use this to re-estimate
the OH distribution near Tethys [2,4]. Earlier we had shown that fast
ion sputtering produces copious O2 at the temperatures expected on
Jupiter's icy satellites [3], and the temperature dependence of the O2
yields could be understood if we assumed that O2 is produced in the ice
and is subsequenty ejected. Others have reported luminescence from O2 in
irradiated ice and recently Spencer et al. [5] report the observation of
O2 in a solid matrix on Ganymedes trailing hemisphere. Therefore, we
have also used data on sputter ejection of molecular species,
reflectance in the UV, and luminescence spectra to determine the
composition of irradiated water ice on the surfaces of the icy

References: [1] Westley M. S. et al. (1994) Nature, submitted. [2]
Johnson R. E. (1993) EOS Trans., 74, 569. [3] Johnson R. E. (1990)
Energetic Charged Particle Interaction in the Atmospheres and Surfaces,
Springer-Verlag, Berlin. [4] Shemansky D. E. et al. (1993) Nature, 363,
329. [5] Spencer J. R. et al. (1994) JGR, submitted.

26.19-P Luhmann J. G.
Extended Atmospheres of the Outer Planets: Potential for
In-situ Detection

Several years ago, Mendillo and coworkers (e.g., see J. Geophys. Res.,
99, p. 8403, 1994 paper by Flynn et al.) discovered the tremendous
extent of the visible Jovian sodium nebula. This disk-shaped neutral
cloud, which has its origins in charge exchange between corotating
magnetospheric ions from Io and the surrounding neutral Io torus,
reaches hundreds of Jovian radii from the planet in the images obtained.
These investigators have modeled the required source strength and find
it to be ~10^26-10^27 sodium ions per second, consistent with
theoretical predictions based on Voyager in-situ measurements. Similar
nebulae are expected to result from charge exchange in the satellite
tori of Saturn, and in other (less visible) elements such as oxygen for
Jupiter. The Ulysses and upcoming WIND and ACE spacecraft ,will have
detectors capable of measuring the singly charged ions produced when
these extended atmospheres are ionized by photoionization, charge
exchange, or impact ionization in the solar wind. We predict the
expected fluxes and distribution functions that should characterize the
"signature" of these most extended planetary atmospheres at distances of
several AU from their sources.

26.20-P Mallama A.
Models for Saturnian Eclipses of Mimas, Enceladus, Tethys,
Dione, Rhea, Titan, Hyperion, Iapetus, and Phoebe

A saturnian eclipse of any satellite presents an opportunity for very
accurate astrometry. The satellite position is determined by fitting the
observed light curve to a model light curve which is based on the
refraction of sunlight in Saturn's atmosphere, the effect of the solar
phase angle, and the satellite's albedo characteristics.

Analysis indicates that the minimum altitude of a refracted ray of
sunlight which has been reduced to half-intensity, ranges from 165 km
above Saturn's 1-bar surface for an eclipse of Mimas, to 360 km for
Phoebe. The solar phase angle can effect the derived posltion of Titan
by almost 100 km, though it is not more than 25 km for the other smaller
satellites. The hemispheric albedo effect is about 200 km for Iapetus
because of the extreme albedo contrast between its leading and trailing
hemispheres, but is 40 km or less for all other satellites.

A computer program developed for jovian eclipses of the Galilean
satellites (Icarus, 92 , 324-331, and Icarus, 97, 298-302) has been
expanded to generate model light curves for the satellites of Saturn.
The time of half-phase can be derived by fitting the observations to the
model light curve. Half-phase is defined to be the moment when a
rerracted half-intcnsity light ray from the center of the Sun's disk
strokes the center of the eclipsed satellite.

26.21-P Denk T. Jaumann R. Neukum G.
A New Search for Surface Features on the Dark Leading Side of

The most unusual property of the Saturnian satellite Iapetus is the
extreme difference in brightness between the leading and trailing/polar
hemispheres (albedo 0.05 and 0.5, respectively). While individual
Voyager images of the bright side show many features like craters or
scarps, dark Cassini regio area appears featureless.

The main reason for this lack is the low signal-to-noise ratio of the
raw-image data. Depending on the used filter and exposure time, the dark
side is (on average) only about 2-3% brighter than the dark sky. Another
problem is the low resolution of the Cassini regio frames (23 km/pixel
or worse).

The S/N ratio can be improved using subpixel registration, and
sumrnation of different frames from one image sequence. This allows S/N
ratio to be improved by a factor of greater than three.

Preliminary results indicate that features appear within Cassini regio
at a higher signal level. One example is a circular structure with a
diameter of about 400 km in the northern part of the center of the
leading side. A description of these features is presented.

26.22-P Pascu D. Rohde J. R. Foechterle L. D.
B, V Photometry of Helene (SXII)

Helene is a Lagrangian librator of Saturn/Dione, leading Dione in its
orbit by 60 degrees. Photometry of Helene is important for explaining
the source of asymmetry in the photometry of Dione since it shares
Dione's orbit but is too small for endogenic processes to affect its
surface composition.

We have observed the Saturnian system over several apparitions and
present here the B, V results for Helene for some of them. Our strategy
was to relate the photometry of Helene to that of the brighter
satellites, especially Titan. Dimensions of Helene obtained by Voyager
were used to make aspect corrections and to compute an albedo.

Our results show Helene to be neutral in color, (B-V)= 0.6, and 0.5 mag.
brighter than earlier estimates. A preliminary light curve indicates
that Helene is fainter at eastern elongation than at western -- opposite
to that for Dione.

26.23-P Zellner B. Seidelmann P. K. Pascu D. Kowal C. Wells E.
Currie D.
Recovery of Inner Satellites of Uranus

None of the ten small inner Uranian satellites Cordelia through Puck
have been seen since Voyager 2 lett the system in January of 1996, and
the current un-certainties in their orbital positions range from 5 to
110 degrees. We expect to recover several of these objects in images
taken with the Planetary Camera of the Hubble Space Telescope in August
of 1994. Exposures will be made at three epochs a few hours apart in B,
V, R, and I filters. Miranda and Ariel will also appear in the frames
and will provide both photometric standards for low-precision
colorimetry of the inner satellites and astrometric benchmarks for
greatly improved knowledge of their orbital motions. The improved
orbital positions should allow spectrophotometry of several of the
brighter objects in subsequent work with HST.

This work is based on observations with the NASA/ESA Hubble Space
Telescope obtained at the Space Telescope Science Institute, which is
operated by the Association of Universities for Research in Astronomy,
Inc., under NASA contract NAS5-26555. This work was supported by STScl
Grant GO-5321.

SESSION 27 ....... Asteroids I
Friday, 8:30 - 10:00 Crystal Ballroom A
C. Coombs and R. Greenberg, Moderators

27.01 Bottke W. F. Jr.* Nolan M. C. Greenberg R. Vickery A. M.
Melosh H. J.
Provenance of the Spacewatch Small Earth-Approaching Asteroids

Recent discoveries of small Earth-approaching asteroids by the
Spacewatch telescope (referred to here as S-SEAs) reveal more bodies
with diameter D < 50 m than had previously been inferred from the
distribution of larger bodies (D > 100 m). Assuming that the observed
S-SEAs represent a steady-state excess of small bodies with perihelia
and aphelia near 1 AU, e <- 0.35, and i between 0 to ~30 degrees, we try
to match S-SEA dynamical and size constraints to a given source by
following the orbital and collisional evolution of test bodies from that
source. We use the Monte-Carlo dynamical evolution code of Melosh and
Tonks, modified to include an impact disruption model (based on a map in
orbital a, e, i space of collision probabilities and mean impact
velocities found using actual main-belt and near-Earth asteroid orbits),
fragmentation after collisional disruption, and observational selection

Our results show that main-belt objects exiting the 3:1 and nu-6
resonances when they become Earth-crossing are unlikely to become
S-SEAs; objects from these sources frequently disrupt or impact the
Earth before reaching the low-e orbits of the S-SEAs. Furthermore,
though frequent impacts with the main-belt would produce fresh
collisional debris, there is no reason for the process to create the
clear excess of small bodies seen as S-SEAs. A stochastic break-up of a
NEA in a low-e orbit could produce the observed population, but (a)
disruption in this region is unlikely, (b) fragments would show
correlated inclinations, and (c) S-SEA spectra does not match observed
NEA spectra. However, planetary ejecta from either the Earth-Moon system
or Venus could produce a small body excess spanning the current S-SEA
orbits within ~10 Myr after ejection. Ejecta from these regions is
collisionally decoupled from the main-belt, allowing it to survive and
maintain its steep size-frequency distribution. Small-body ejecta from
Mars does not appear to dynamically evolve into the low-orbits necessary
to match S-SEA dynamical constraints.

27.02 Whiteley R. J.* Tholen D. J.
CCD Search for Lagrangian Satellites of the Earth

The Lagrangian L4/L5 points of the Earth were imaged using the
University of Hawaii 2.2m telescope, with the Tek2048 detector, on May
5-7,1994, and will be again on July 6-8. The objective is to locate
sub-kilometer sized asteroids orbiting at Earth's libration points, or
failing this, to set better upper limits on the distribution of such a
population. Previous searches of these areas have used various
photographic plate techniques, but no asteroids have ever been
discovered at the L4/L5 points of the sun-earth system.

There are many observational constraints to this kind of search. All the
observations, by definition, are at airmasses >1.5. The objects in
question move roughly 150"/hour in right ascension, so discovery and
recovery of them has to happen in essentially real time. Tme is critical
for another reason: the Lagrange clouds are only visible for about 2.5
hours on a given night. Fortunately, the skies above Mauna Kea are
extremely dark, the seeing is routinely sub-arcsecond, and the 2.2m
telescope and Tek2048 have been used successfully many times to search
for moving objects at this level of sensitivity.

We hoped to improve on earlier work by ~3 visual magnitudes, and from
our initial data it appears as though the limiting sensitivity for the
initial search was R~23. This corresponds to asteroids in the 200-300m
size range. Preliminary results show that no 3 sigma detections were
recorded, although there may be one candidate object at the 2 sigma
level. The planned July observations should provide better constraints
on existence of Lagrangian satellites of the Earth.

27.03 McCord T. B.* Coombs C. R. Morris J. Pershing D.
Remote Observations of Meteoroid Events on Earth

On February 1, 1994, two U.S. defense satellites detected an energy
burst over the central Pacific Ocean. Here, we present our analysis of
this event made from original and unprocessed data. We estimate the
energy released in this burst to be ~63 KT of TNT and conclude that its
source was a meteoroid entry into Earth's atmosphere. In the first few
seconds following atmospheric entry, the meteoroid was observed to brake
into two clusters and several smaller fragments and form debris clouds
which lasted for nearly 20 minutes as they drifted westward at a rate of
~70km/hr. Assuming a meteoroid entry velocity of 10-30 km/s, a mass
range of 0.6-5 x 10^6 kg is derived. Further, if spherical and composed
of iron, the meteoroid would have been ~5.6-10.7 m in diameter. If
chondritic in composition, the meteoroid would have been slightly larger
at 6.8-14 m in diameter. Based on the debris cloud formation and
dissipation, the meteoroid was most likely silicate, or at least
non-iron in composition.

27.04 Meisel D. D.* Getman V. S. Mathews J. D. Morton Y. T.
Zhou Q. Roper R. G. Marsden B. G.
An Unusual Bolide with a Near-Earth Orbit

On April 8, 1989 at 0526 UT, a very bright (-10 visual magnitude),
slowly moving bolide (4 second duration) was observed from three
locations in Puerto Rico. It was detected by instruments in place as
part of the AIDA '89 (Arecibo Initiative for Dynamics of the Atmosphere)
campaign-an All-Sky camera, an intensified video camera system, a CCD
Infrared Imager, an imaging narrow band spectrometer (MORTI), and a IDI
(Imaging Doppler Interferometer) radar. Our calculations indicate that
approximately 140 kg of original (photometric) mass entered the
atmosphere on a steep trajectory, about 32 degrees from vertical, and
was luminous from 65 km altitude at 15 km/s velocity down to 25.6 km
altitude and 1-2 km / s velocity. The apparent radiant (alpha = 190.8
degrees delta = -7.7 degrees, 1950), and initial velocity (15.5 +/- 0.3
km/sec ) shows that the object was able to make several close approaches
to the earth at both ascending and descending nodes before final entry.
A stony composition was indicated in a number of independent ways
including the light curve and ablation calculations, but the
spectroscopic evidence is the strongest. Observations in the near IR
(868 nm) show only weak iron emission, although the excitation
temperature of the iron vapor was determined to be 3100 K. Ionized
calcium which is so strong in most meteor spectra and should have been
seen in the same spectral region, could have been at most 10% of the Fe
intensity. Non-detection of sodium by lidar measurements limits its
contribution to less than 0.1% by mass. Thus enstatite seems to be the
most plausible mineral constituent for the meteoroid and a connection
with aubrite meteorites is inferred. Extensive fragmentation during
descent, end height above 20 km, and the rugged terrain of the predicted
impact area seems to preclude any recovery of significant meteorite

27.05 Holman M.* Wisdom J.
Meteorite Delivery from the 3:1 Kirkwood Gap

That small bodies in the family of chaotic zones associated with the 3:1
mean motion resonance with Jupiter can undergo intermittent jumps in
eccentricity is now well-established (Wisdom 1982, 1983, 1985). These
excursions in eccentricity typically peak near 0.6. Such eccentricities
permit objects in the 3:1 Kirkwood gap to cross the orbit of Eartb in
relatively short times, much shorter than the collisional lifetime of
meteorites inferred from cosmic ray exposure ages (10^7-10^8 years).
However, recent theoretical research into the dynamics near mean motion
resonances reveals even more surprising behavior. Studies of the
averaged elliptic restricted three-body problem show that chaotic zones
associated with some mean motion resonances can lead to eccentricities
near unity (Yoshikawa 1989; Klafke, Ferraz-Mello, and Michtenchenko
1992). The long-term behavior at the 3:1 Kirkwood gap is studied with a
full dynamical model that includes the gravitational perturbations of
all the planets without averaging. It is observed that some test
asteroids alternate between a mode where the eccentricity is primarily
low (0.1-0.2) with jumps in eccentricity (0.5-0.7) and a mode where the
eccentricity ranges between 0.7 and nearly 1. The effects of this high
eccentricity behavior on the timescale and efficiency with which small
bodies are delivered from the regil)ns near the 3:1 Kirkwood gap to
Earth-crossing orbit are examined.

This work was supported in part by the NASA Planetary Geology and
Geophysics program under grant NAGW-706, by an NSF Presidential Young
Investigator Award AST-8857365, and by a post-doctoral fellowship from
the Japan Society for the Promotion of Science.

27.06 Nolan M. C.* Greenberg R.
Asteroid Family Formation

Asteroid families are probably the daughter products of disruptions of
larger parent asteroids, whose fragments have not migrated very far in
orbital space so that their proper elements remain similar. The orbital
and spectral similarity within some families indicates that their
members are genetically related. The apparent velocity dispersions among
family members (50-200ms^-1) are a few times larger than those found by
impact experiments or by our numerical simulations. This discrepancy
indicates that there may be a scaling problem in comparing experimental
results with the asteroid belt. One explanation may be the errors in the
determination of velocity dispersions of families [e.g., Carpino et al.
Icarus 1986].

We consider an alternative explanation: Special impacts may be required
to create families. Based on hydrocode modeling of impact mechanics and
models of asteroid populations, we find that most asteroids large enough
to form observable families had earlier been fragmented (but not
disrupted) by repeated impacts. Thus, unless the initial fragments are
re-cemented, family-producing collisions may be atypical. This
conclusion would resolve a dilemma posed by the impact record on the
asteroid 243 Ida. Preliminary results from the Galileo spacecraft
[Belton et al. DPS 1993] suggest that Ida, and thus the Koronis family
to which it belongs, may be over a billion years old. That value is
consistent with Binzel's [Icarus 1988] estimate of less than 2-4 Gy. If
families are that long lived, one would expect to find many more of them
than we observe [Harris DPS 1993], unless family formation occurs only
under special circumstances, such as the re-cementing mentioned above.
Another special circumstance might be low-velocity collisions (V <~ 1km
s^-1) which are quite probable [Bottke GRL 1994], and yield larger
fragments, and partition a larger fraction of the impact energy into
kinetic energy of fragments than do higher-velocity impacts. Ongoing
simulations of a wide variety of collision scenarios, are helping
constrain family-producing conditions.

27.07 Harris A. W.*
Tumbling Asteroids 2: 3188 Seleucus

Small asteroids which are spinning very slowly may have a timescale of
wobble damping which is long compared to their collisional lifetime, or
even the age of the solar system (J. A. Burns and V. S. Safronov, Mon.
Not. R. Astron. Soc. 165, 403-411, 1973), thus a highly irregular body
may be in a state of "tumbling" motion (nonprincipal axis rotation),
similar to that of the nucleus of P/Halley, and thus have a complex
lightcurve that is not a simple periodic function (A. W. Harris, Icarus
107, 209-211, 1994). The asteroid 3288 Seleucus (1982 DV) appears to be
an example of this phenomenon. Seven nights of data, taken in 1982, have
been analyzed, and no unique period can be found to fit the data. Below
is a plot of the noise spectrum of the data, showing that no single
penod fits the data satisfactonly, and there appear to be two or more
non-harmonic frequencies present. I have constructed best-fit composites
for each of the periods labeled below, and verified that none of them
adequately represents the data. Thus I conclude that Seleucus is an
example of a "tumbling" asteroid.

This work was supported at JPL/Caltech under contract from NASA.

27.08 Ostro S. J.* Choate D. Dendrenos P. Franck C. R. Giorgini J.
Hills D. L. Howard D. Jurgens R. F. Keesey M. S.
Mitchell D. L. Rose R. Rosema K. D. Slade M. A.
Strobert D. R. Winkler R. Yeomans D. K.
Asteroid Radar Astronomy at Goldstone in the 1990s

Between January 1990 and June 1994 the Goldstone Lambda 3.5-cm
(8510-MHz) radar detected 12 asteroids: 4 Vesta, 7 Iris, 97 Klotho, 194
Prokne, 324 Bamberga, 1981 Midas, 3103 Eger, 4179 Toutatis, 4953 (1990
MU), 5189 (1990 UQ), 1990 MF, and 1990 OS. Echo bandwidths place the
following lower baunds an D/P, where D (km) is the maximum pole-on
breadth and P (h) is the apparent rotation period: 0.50 for Midas, 1.2
for 1990 MU, 0.13 for 1990 OS, and 0.10 for 1990 UQ. For each target,
estirnates of radar cross section and circular polarization ratio
constrain near-surface bulk density and roughness.

Since 1990 much effort has been devoted to optimizing Goldstone's
capability of studying near-Earth objects (NEOs). Three different
configurations are now available. The nominal system on the 70-m antenna
(DSS 14) uses separate feed horns for transmitting and receiving, and
the >-15 s required for TX/RCV switching renders the system useless for
targets much closer than ~0.025 AU (~10 lunar distances, ~25 s of echo
time delay) and unwieldy even for targets twice that far. However, an
additional, recently installed, single-horn system (first used in 1994,
for 1990 MU) shrinks the switching tirne to one second, letting DSS 14
observe NEOs as close as a few lunar distances. Closer targets will
require two-station observations, with DSS 14 transmitting continuously
while DSS 13, a 34-m antenna 22 km away, receives. That configuration
(used in 1992 for Toutatis) will also be required for delay-Doppler
imaging that places at least 100 frequency cells on any NEO for which
the echo's roundtrip time delay RTT (s) satisfies RTT < P/D.

Goldstone is scheduled to observe seven asteroids during the next two
years. Delay-Doppler imaging experiments, expected to resolve the target
into >-100 cells, are planned for 1620 Geographos in August 1994 and
1991 JX in June 1995. A new data-acquisition system to be implemented by
1996 will be entirely digital after the receiver's low-noise front end,
and will attempt to adopt Arecibo's user interface and data-handling

After completion of the Arecibo upgrade, Goldstone will play an
important complementary role to that much more sensitive instrument.
Arecibo will see twice as far and access three times as much volume as
Goldstone, but Goldstone's fully steerable antennas will have
solid-angle and hour-angle windows equal to two and three times those of
Arecibo. Goldstone will extend the rotational and orbital phase
coverages for targets visible at both sites and will help to secure the
arbits of NEOs discovered after their exit from Arecibo's declination

27.09 Mitchell D. L.* Ostro S. J. Rosema K. D. Campbell D. B.
Chandler J. F. Shapiro I. I. Hudson R. S.
Large-Scale Topography on Main-Belt Asteroids: Evidence from
Arecibo Radar Spectra

Arecibo lambda 13-cm radar spectra obtained between 1980 and 1989 of the
main-belt asteroids 7 Iris, 9 Metis, 12 Vlctoria, 216 Kleopatra, and 654
Zelinda exhibit evidence for large-scale topography. These asteroids
range in diameter from 113 to 200 km and include members of the S, C.
and M classes. Radar spikes appear in narrow ranges of asteroid rotation
phase in spectra of Iris, Metis, and Zelinda. These "glints," which can
briefly account for a significant fraction of the total echo power, are
likely caused by single back-reactions from large, flat regions. For
Iris (left figure), the glint is resolved in Doppler frequency, allowing
us to constrain one dimension of the flat region to be >~20 km in
extent, or roughly 10% at the asteroid's diameter.

Radar spectra of Victoria and Kleopatra are bimodal over >30 degrees
ranges of asteroid rotation phase, indicating non-convex shapes. Echoes
from Vlctoria are strongly bimodal near one rotation phase (right
figure) but not near rotation phases ~180 degrees away, suggesting an
irregular, non-convex shape. The upgraded Arecibo is expected to provide
a >10-fold increase in radar sensitivity that should permit
delay-Doppler imaging of Iris in late 1995, Metis in 1997, Kleopatra in
1999, Zelinda in 2002, and Victoria in 2007.

Figure 1, which appears here in the hard copy shows, The figures plot
radar cross section (km2) per frequency resolution element (90 Hz for
Iris and 30 Hz for Victoria) vs. Doppler frequency (Hz). The vertical
bar at the origin shows +-l standard deviation of the noise. The spectra
may be thought of as one-dimensional images in which each frequency
"pixel" measures the echo power in a strip parallel to the asteroid's
apparent rotation axis projected onto the sky plane.

27.10-T Durda D. D. Jayaraman S.
On the Origin and Dynamical Lifetime of the Near-Earth Asteroid

Recently a number of small (lOm scale) asteroids in low-eccentricity and
near-Earth orbits have been discovered by the Spacewatch program
(Rabinowitz et al., Nature 1993.). Their numbers appear to exceed that
expected after accounting for observational selection effects,
suggesting the existence of a previously undetected near-Earth asteroid
belt. More-over, a significant fraction of the small objects are in
corotational resonance with the Earth. We examine through numerical
integrations, the possibility that at least some of these objects may
have been delivered to their present orbits by Yarkovsky's effect, a
non-gravitational perturbation due to the reaction force resulting from
the angular displacement of the reemitted thermal radiation from a
rotating body. For retrograde rotators the Yarkovsky effect may cause
the orbital collapse of asteroids in the lOm size range in a manner
analogous to the Poynting-Robertson light drag on micron sized dust
particles. The puported asteroid belt may then be similar to the
recently discovered asteroidal dust ring in resonant lock with Earth
(Dermott et al, Nature 1994).

The eccentricities of these asteroids provide a clue to how long they
have been in resonance. We also calculate the dynamical lifetime of such
asteroids before they are perturbed out of their Earth-like orbits due
to a close encounter with the planet and therefore determine the flux of
such asteroids needed to maintain the observed distribution.

27.11-P Bowell E. Muinonen K. Harris A. W. Howell S. B.
A Comparison of Current and Planned NEO Detection Systems

We compare the performance characteristics of three ongoing and six
planned telescope/detector systems designed to discover near-Earth
objects (NEOs, both asteroids and comets). Our reference is the modeled
performance of the Spaceguard NEO Survey, in which six 2.5-m telescopes
scan the sky to limiting magnitude V(sub)lim, = 22 for at least a
decade. Over a 25-yr span, the Spaceguard system appears capable of
discovering 88% of Earth-crossing asteroids (ECAs) and 43% of
Earth-crossing comets (zero warning time) larger than 1 km diameter
Because of modeling difficulties for comets, we consider only ECA
detection. To make a reasonably fair comparison; we assume that
detection of putative ECAs is made solely on the basis of their
sky-plane motions being markedly different from those of main-belt
asteroids. Thus, we assume that follow-up and orbit-determination of
putative ECAs is accomplished separately from the main survey effort.
Such a modus operandi resembles that of the University of Arizona's
Spacewatch system; but it is not like that of Spaceguard, in which, much
of the time, a fixed area of the sky is scanned repetitively and
follow-up is largely accomplished automatically. We further assume that
each system has only one telescope. We have used A. W. Harris's method
of performance characterization (,4stron. Soc. Pacific Conf. ,Ser., in
press), bearing in mind that the method is very approximate and is poor
for very broad sky coverage. We have also used a 10^5-member model ECA
population to predict discovery rates as functions of diameter and
survey duration (cf E. Bowell and K. Muinonen 19941 in press in Hazards
Due to Comets and Asteroids). We comment on the strong and weak features
of the nine systems, we present figures of merit based on the modeled
discovery rates of ECAs larger than a given diameter, and we estimate
the smallest ECA that can be detected with reasonable certainty before
it is likely to collide with the Earth.

27.12-P Scheeres D. J. Ostro S. J. Hudson R. S. Werner R. A.
Orbits About Asteroid 4769 Castalia

We have modeled the gravity field near this km-sized Earth-crosser by
combining a radar-derived 3-D shape model (Hudson & Ostro 1994, Science
263, 940), an assumed uniform density of 3.5 g/cc, and an assumcd 4.07-h
rotation about the model's largest moment of inertia. That field, which
can be expressed in closed form as a polyhedron or in terms of standard
gravitational harrnonic coefficients, is the basis for exploring the
stability of orbits near Castalia and for examining trajectories of
impact ejecta. A Jacobi Integral is defined for this problem and has
been used to generate zero-velocity surfaces for the asteroid.

Four synchronous circular orbits, all close to the equatorial plane and
separated by roughly 90 degrees in longitude, exist but are unstable.
Therefore the model exhibits Type II gravitational dynamics about it
(Scheeres 1994, Dynamics about uniformly rotating triaxial ellipsoids,
Icarus 110, in press): direct orbits within three to four mean radii are
unstable and either crash onto the surface or escape. Stable orbits have
been found that come within meters of the surface and persist
indefinitely under the gravitational attraction of Castalia alone. All
these stable orbits are retrograde and close to the equatorial plane.
The minimum periapsis of stable orbits increases as the inclination
decreases from 180 degrees to 0 degrees. All direct, nearly circular,
equatorial orbits within 1.75 km of the asteroid's center of mass appear
to be unstable and either escape within days or crash onto the
asteroid's surface.

Any rock ejected from an impact event will either return to the surface,
enter a stable orbit about the asteroid, or escape. The actual
trajectory will depend on surface location and initial speed and
direction. For any point, we can calculate the speed upsilon(sub)R,
below which ejecta definitely will return to the asteroid, regardless of
launch direction; and the speed upsilon (sub)E, above which ehcta will
definitely escape, regardless of launch direction. Both upsilon (sub)R
and upsilon (sub)E vary over Castalia's surface, as does the interval
upsilon (sub)D = upsilon (sub)E - upsilon (sub)R. upsilon (sub)R >- 0.08
m/s everywhere and is as high as 0.4 m/s in some locations. upsilon
(sub)E <- 1.05 m/s everywhere and is as low as 0.8 m/s in some
locations. upsilon(sub)D<- 0.9 m/s everywhere. Additional calculations
are expected to refine these constraints.

There is no clear transport mechanism between the asteroid's surface and
stable, direct, nearly equatorial orbits. Such a mechanism would require
either a chaotic path about the asteroid (which would have a low
probability of occurring and a high probability of crashing or escaping)
or third-body perturbation forces. However, it is possible for some
ejecta to be launched into stable orbits. The most likely such orbits
are retrograde with respect to the asteroid's rotation pole, the ejecta
emanating from the vicinity of the longer ends of the asteroid with
speeds in the range of 0.9 m/s. These orbits can come arbitrarily close
to the long ends af the asteroid.

27.13-P Campbell D. B. Black G. J. Ostro S. J.
Asteroids and Comets: Future Imaging Opportunities with
Earth-Based Radar Systems

Major improvements currently underway to the Arecibo 305m antenna and
Lambda 13cm radar system will increase the radar system's sensitivity by
a factor of about 20. Recent upgrades to the Goldstone Lambda 3.5cm
radar system have also improved its sensitivity. While the Arecibo
system will have significantly greater sensitivity than Goldstone, the
Arecibo antenna's limited declination coverage of -2 degrees to +38
degrees means that the Goldstone system will be important for the
investigation of near-Earth objects with declinations outside this
range. Over one hundred mainbelt and near-Earth asteroids per year will
be observable with the new Arecibo system, and there will be a number of
good imaging opportunities each year. While only a few imaging
opportunities are predicted for comets over the next 10 years, the
system will be able to respond to unexpected cometary opportunities such
as that for Comet IRAS-Araki-Alcock, which approached to within 0.03 AU
of the Earth in 1983. For sufficiently close objects, both the Arecibo
and Goldstone systems are capable of delay-Doppler imaging at 20m
resolution. However, the ambiguities inherent in the delay-Doppler
technique can make these images difficult to interpret, and new
approaches are needed. We are currently investigating the applicability
of VLBI techniques to the direct synthesis of images of asteroids and
comets illuminated by a radar signal. For example, at Lambda 3.5cm, the
newly-completed Very Long Baseline Array (VLBA) will have approximately
50m resolution at the distance of Geographos during its close approach
to the Earth in late August 1994. While the sensitivity of
Goldstone-VLBA will not sustain 50m resolution and problems related to
sidelobe response may be difficult to solve, the resultant image would
be several pixels across and should provide complementary information to
that from delay-Doppler imaging. A similar opportunity for
Goldstone-VLBA observations of 1991 JX will occur in June 1995.

27.14-P Melosh H. J. Bottke W. F. Jr.
Binary Asteroids and the Formation of Doublet Craters

Out of the approximately 28 known terrestrial impact craters with
diameters larger than 20 km, at least 3 are doublets created by the
nearly simultaneous impact of objects of comparable size (Melosh and
Stansberry, 1991, Icarus, 94, 171.). Recent radar observations of
Earth-crossing binary asteroids (e.g. 4769 Castalia and 4179 Toutatis)
by Ostro et al. (1990, Science 248, 1523.; 1993, BAAS, 25, 1126.)
suggest that these bodies could produce doublet craters if their
components separate before impact. Despite this, proposed mechanisms,
such as tidal disruption during the binary asteroid's final approach and
break-up due to atmospheric friction, are insufficient to produce
significant separation of the components. However, Farinella (1992,
Icarus, 96, 284.) and Chauvineau et al. (1994, to be submitted) have
shown analytically that binary asteroids encountering the Earth can have
their orbital energy significantly modified by tidal perturbations,
causing the components to (a) collide, (b) escape "to infinity", or (c)
go into orbit around one another. Endstate (c) is directly applicable to
doublet craters production, since the components are likely to be near
their furthest extension (i.e. aphelion) at impact.

To quantify this process, we developed a model which numerically
integrates close encounters between binary asteroids (with random
initial orientations) and the Earth. We find that loosely bound contact
binaries encountering the Earth at distances less < 5 R(sub)Earth can
become well separated, confirming the general results of Chauvineau et
al. (1994). In addition, we find that the distance between these
components can increase with additional planetary encounters.

Finally, to verify that the binary asteroids/tidal forces can produce
doublet craters, we combine this model described above with the
Monte-Carlo dynamical evolution code used by Bottke et al. (1994, BAAS,
this issue), which simulates the orbital evolution of asteroids in the
terrestrial planet region. We find that our results are consistent with
Earth's cratering record; ~15% of the binary asteroids impacting the
Earth are separated by 4 times (or more) of their mean diameters. These
results imply that many large NEAs are/evolve into binary asteroids.

27.15-P Farinella P. Davis D. R.
Why Are Small Earth-Approachers Different?

Rabinowitz (1993,Astrophys. J. 407, 412) and Rabinowitz et al (1993,
Nature 363, 704) have recently shown that small NEAs (diameter D = 10 -
100 m) have an orbital distribution very different from that of NEAs
with D > 1 km. Small NEAs often have semimajor axes close to 1 AU and
low eccentricities, a combination quite rare among larger objects. We
propose that, rather than being due to a different source for these
bodies (e.g., the Moon), this difference may be related to the size and
orbit dependence of their average lifetime tau (sub) coll vs.
collisional disruption by other asteroids (both main-belt ones and

Assume that the number N of NEAs of a given size is given by the simple
evolution equation:

Equation 1

where F is the flux of new objects coming from the main asteroid belt
through resonant routes, and tau (sub) dyn ~ 10(sub)^8 is the average
lifetime vs. hyperbolic ejection and impact with the Sun or the planets,
independently of size. Then the overall average lifetime of NEAs is tau
= (tau = (tau ^-1(sub)coll + tau^-1(sub)) dyn ^-1, and the steady-state
population is just F(sub)tau, namely proportional to the lifetime. Now
assume that tau (coll)~(10^8 yr) X (D/1 km)^1/2 for bodies having part
of their orbit in the main asteroid belt (say, aphelion distance Q >1.7
AU), and tau (sub)coll ~(5 x 10^8 yr) X (D/1 km)^l\2 for Q well inside
the inner edge of the main belt. This is based on a typical collisional
lifetime of several times 10^8 yr for main-belt Gaspra-sized (D~10 km)
objects, the dependence tau (sub) is variable to D^1\2 as predicted by
Dohnanyi's equilibrium size distribution for a population of
collisionally interacting asteroids, and a flux of impact energy
enhanced by a factor ~50 in the main belt with respect to Earth-like

As a consequence of the above relationships, it is easy to see that tau
increases only by a factor 2 (from 50 to 100 Myr) for km-sized NEAs, but
by about a factor 10 (from ~10 to 100 Myr) for small NEAs, when the
orbits go from Q >1.7 AU to Q <1.7 AU. Therefore, small objects should
become more abundant relative to larger ones bsr a factor ~5 when the
orbits have semimajor axes and eccentricities leading to aphelia well
inside 1.7 AU. Note that this mechanism predicts that the size
distribution of NEAs should not depend on the orbital inclination, as
observed. We are currently working to a more refined model of NEA
demography, making use of the two-components capability of the
collisional evolution code described in Davis et al. (1989, Asteroids
II, 805-826).

27.16-P Hills J. G. Solem J. C.
Shaping of Near-Earth Asteroids by Tidal Forces

The Near-Earth objects that have been observed by radar to date have
conspicuously elliptical shapes. If they are "rubble piles" rather than
single chunks of competent material, then their shapes might be the
result of close gravitational tidal encounters with Earth or other
planets. NEO's are removed from the population by hitting Earth or
suffering encounters with Earth that pump up their orbital energies, so
their orbits become very strongly perturbed by Jupiter. The cross
section for tidal encounters is larger than that for direct collision
with Earth, so a sizable fraction of the NEO's have passed inside the
tidal radius of Earth. We model the asteroids as agglomerations of
spherical rocks, bound together only by gravitation. We used an N-Body
code to determine the amount of elongation they acquire in a tidal
encounter as a function of impact parameter and velocity with respect to
Earth. We assume that the component,of the velocity in the radial
direction between two colliding spheres is damped out in any collision
between them, so the corresponding kinetic energy is converted to heat.
The model exhibits a remarkable scaling relation that makes detailed
geometric and energetic features of the tidal distortion independent of
asteroid size. It is found that a tidal encounter causes the object to
become less gravitationally self-bound and to become permanently
distorted into a cigar-shape. Tidal encounters appear to be a likely
explanation for the elongated structure of Earth-Crossing asteroids.
This work was supported by the Department of Energy.

27.17-P Vasilyev S. V.
New Determinations of Asteroid's Polarimetric Parameters

The method on the basis of the principal component analysis had been
applied for asteroid polarization phase curves. It has allowed to
approximate the negative branches and determine the polarimetric
parameters. The polynomial and trigonometric approximations applied
usually require full measured curves or observations at quite specific
phase angles and reflect their peculiarities differently. The proposed
method is used to automate and improve the interpolation and
extrapolation of data. The table contains the polarimetric parameters of
some asteroids estimated for the first time or specified by principal
component method using available data including results of out

The possibilities to optimize the planning of asteroid polarimetric
observations using the principal component method are discussed.

SESSION 28 ....... Pluto
Friday, 8:30 - 10:00 Crystal Ballroom B
R. H. Brown and J. A. Stansberry, Moderators

28.01 Ip W.-H.* Lara L.
Pluto's Upper Ionosphere: A Quick Look

Since Pluto is not likely to have an intrinsic magnetic field, its
interaction with the solar wind flow could take the form of either
cometary-like interaction in case the atmospheric outflow is
sufficiently strong or Mars/Venus-like interaction in the other extreme.
Under both conditions, the ionospheric structure and composition are
expected to be significantly modified by the plasma convection and
diffusion effects. In this study we consider the location of the
ionopause where the external magnetic pressure is balanced by the
ionospheric plasma pressure inside this boundary. Photochemial models
will be computed to examine the possible ion composition distributions
in different regions of Pluto's ionosphere.

28.02 Young L. A.* Elliot J. L. Tokunaga A. Owen T. Cruikshank D.
de Bergh C.
Spectroscopic Observations of Atmospheric CH4 on Pluto

High-resolution spectra of Pluto and Charon were taken with the CSHELL
eschelle spectrograph at the IRTF on 25-26 May 1992, to find the column
height of gaseous methane in Pluto's atmosphere. The spectral resolution
lamda/delta lamda was 13,300 and the spectrum spanned 1.6618 -1.6670
micrometers. The spectral range included the R(0) line and the Q branch
of the 2v(sub)3 band of methane. At this resolution, the narrow
vibration-rotation lines are readily distinguishable from the broad
absorption of the methane frost.

The methane absorption was calculated line-by-line on a fine array for
all 65 methane lines listed in the HITRAN database for this spectral
range. The solar absorption was supplied by the KPNO Solar Atlas, and
the telluric absorption was calculated with the ATRAN program (Lord
1992, NASA TM-103957). An asteroid and a solar analog verified the
accuracy and applicability of the modeled solar and telluric

Absorption features at 1.6655, 1.6660, and 1.6662 micrometers are too
deep to be explained by telluric absorption. These are interpreted as
being caused by Pluto's atmospheric methane, with a column height of at
least 0.3 cm-A. Whether this abundance is high enough to cause the
heating in Pluto's atmosphere at the 1 microbar level depends on the
distribution of the methane and the total column height of Pluto's

This work was supported, in part, by NASA Grant NAGW-3296.

28.03 Lara L.* Ip W.-H.
Photochemical Modelling of Pluto's Atmosphere

Recent ground-based measurements of Pluto's surface thermal emission are
indicative of a range of 30-40 degrees K for its surface temperature
with 36 degrees K as the most likely value. This new finding reinforces
the idea that just like Triton, Pluto's atmosphere should be
N(sub)2-rich with CH(sub)4 and CO as minor components. We have therefore
constructed photochemical models using different thermal profiles and
eddy diffusion coefficients to explore the possible similarities and
differences between the present atmospheres of Triton and Pluto. The
time variation of Pluto's atmosphere as a function of orbital distance
from the sun will be considered.

28.04 Buie M. W.* Wasserman L. H. Tholen D. J.
The Separate Lightcurves of Pluto and Charon from Hubble Space
Telescope Imaging

Observations of the Pluto-Charon system were made with the Planetary
Camera between 1992 May 21 and 1993 August 18. The images contain 30
rotationally distinct longitudes of the system with 2 images at each
time. Despite the spherical abberation, separate photometric extractions
were possible even at their minimum separation of 0.2 arcsec.

The photometric stability of HST is not very well calibrated for the
level of precision required by these observations. To extract an
estimate for the absolute flux of each body a model magnitude was
computed using the albedo model of Buie et al., Icarus, 97, 211 (1992).
The HST photometry provides a flux ratio between Pluto and Charon.
Applying this ratio to the model system magnitude we infer individual
lightcurves of Pluto and Charon.

Initial results from this work do not reveal any lightcurve variations
for Charon. The mean opposition magnitude, corrected to zero phase, for
Charon is V = 17.33 +/- 0.01 while Pluto ranges from V = 15.21 at
maximum light to V = 15.56 at minimum light. The standard deviation of
the Charon photometry is 0.07 mag. Binning the photometry into 40 degree
bins helps reduced the variation in the Charon photometry to 0.04 mag.
When plotted against sample lightcurves, a plausible upper limit on the
lightcurve amplitude is 0.08 mag. With more detailed statistical
modeling and expected improvements in the image restoration, the upper
limit on the Charon lightcurve amplitude should be reduced even further.

We will present out latest results from the photometric analyses of
these data and their implications for future improvments in mapping the
surfaces of Pluto and Charon. This work was supported by STScI grant

28.05 Trafton L. M.* Stern S. A.
HST Observations of Pluto's UV Light Curve and Charon's UV

Pluto's W light curve is of interest for assessing the haze content of
Pluto's atmosphere and for establishing the relative contributions of
Pluto and Charon to the phase variation of the system observed at longer
wavelengths from groundbased observatories. UV spectra of Pluto were
obtained at seven rotational phases near quadrature on 15 and 17 August
1992 and 19 July-28 August 1993 with the HST using the FOS spectrograph
with the 0.5" aperture to exclude most of Charon's spectrum. The spectra
spanned the interval from 1800 angstrom to 4800 angstrom covered by
three gratings. In addition, Charon's spectrum was obtained at opposite
elongations through the 0.5" aperture in the interval 2300-3200
angstrom. The contamination of Pluto's spectrum by Charon due to HST's
broadened PSF is negligible in most cases owing to Charon's relative
faintness and position outside the aperture. We present Pluto's UV
spectrum and light curve and compare Pluto's light curve with the light
curve of the system at blue wavelengths to show that most or all of the
observed variation is due to Pluto, not Charon. However, Charon could
still have a blue light curve variation as great as 12%. Summation of
all Pluto spectra with solar features divided out are also presented in
search of new spectral features from Pluto's atmosphere or surface. The
results are compared with HST spectra of Triton. The contamination of
Charon's spectrum by Pluto is significant owing to HST's broadened PSF.
Observations of a solar type star were obtained at various distances
from the star with the same instrument to calibrate the scattering due
to the PSF. The results were used to derive Charon's spectrum from the
contaminated spectrum. We also present Charon's UV spectrum and compare
it with Triton's spectrum.

28.06 Roush T. L.* Pollack J. B. Cruikshank D. P. Young E. F.
Bartholomew M. J.
The Geometric Albedo of Charon

Pluto's surface appears to be dominated by solid N2 with minor amounts
of CH4 and CO [1]. The surface of it's moon Charon appears to be
dominated by H20 ice [2,3], although significant amounts of other ices
may be present [4]. In an effort to place greater quantitative
constraints on the surface composition of Charon we have derived it's
geometric albedo (p).

We rely upon previous observations of Charon [2] and solar analog stars
[5,6]. SAO120599 (standard used by [2]) has a visual magnitude of 8.414
+/- 0.031 [7], using color values of other GOV stars 5,6], we find it's
H and K magnitudes to be 7.03 +/- 0.08, and 6.97 +/- 0.07, respectively.
We calculated the flux of Pluto and Charon relative to the sun using the
differential magnitudes of [2], the determined H value for lambda's <
1.9 micrometers and the K value for 1.9 micrometers < lambda < 2.51
micrometers, and solar magnitudes from [6]. We calculated p, correcting
for heliocentric and geocentric distances and using 1164 +/- 22.9,km and
621 +/- 20.6 km as the radii of Pluto and Charon, respectively [8].

At lambda's near 1.5, 1.68, 2.03, and 2.35 micrometers, p was ~0.69,
0.44, 0.61, and 0.21 for Pluto and 0.21, 0.27, 0.13, and 0.22 for
Charon. The Pluto values are lower than those of [9], except at 2.35 ?m
where they are equivalent. The Charon values are equivalent to those of
[9], except at ~1.5 micrometers where our value is lower. At these
lambda's, Pluto/Charon ratios are ~3.28-0.95, spanning the range,
1.15-1.6, exhibited in the visual [10]. In the K photometric filter
( 2.2 +/- .3 micrometers) the average p is 0.40 +/- 0.17 and 0.22 +/-
0.04 for Pluto and Charon, respectively, at this light curve phase (phi
= 0.75). The Charon value is equivalent to that of [11] for phi = 0.06
and greater than that for phi = 0.42. The Pluto value remains higher
than that of [11] for either value of phi.
Research supported by NASA Planetary Geology & Geophysics RTOP
151-01-60-01 and Planetaly Astronomy RTOP 19641-67-03.

References: [1] Owen et al. Science 261 745, 1993 t2] Buie et al.
Nature, 329, 522, 1987 [3] Marcialis et al. Nature 237 1349, 1987 [4]
Roush Icarus 108, in press, 1994 [51 Degewij el al. Icarus 44 541, 1980
[61 Campins e? al. AJ 90 896, 1985 [7? Turon et al. The Hipparcos Input
Catalog, ESA SP-1136, ESA Publ. Div., ESTEC, Noordwijk, Netherlands,
1992 [8] Young and Binzel, Icarr?s 108 in press, 1994 [9] Marcialis et
al. Astron. J. 103 1389 1992 [10] Tholen and Buie AJ 95 1977 1988 [11]
Bosh et al. Icarus 95 319 1992.

28.07 Null G. W.* Owen W. M. Jr.
Pluto/Charon Mass Ratio Determined from HST Observations in

We have analyzed Hubble Space Telescope WF/PC I images of Pluto, Charon,
and a field star, acquired on seven HST visits over a 3.1-day span in
August 1993, to observe Pluto's motion about the Pluto system barycenter
and to determine the individual masses and densities of Pluto and
Charon. These images enabled a significant improvement of our earlier
solution (Null, Owen, and Synnott, AJ 105, 2319,1993) which used a
similar data set obtained in 1991.

The camera distortion model was improved using 14 frames of star cluster
NGC 1850 taken between 1990 and 1993. These frames were taken at
different telescope roll angles, enabling us to determine a significant
(1 part/4000), previously unmodeled non-orthogonality between the row
and column axes of the CCD chip. Distortion stability was also verified.

Analyses of the 1991 and 1993 Pluto data sets, both individually and
together, gave values for the Charon/Pluto mass ratio which are in good
agreement. These values are significantly higher than our previous
mass-ratio solution of 0.0837 and are significantly lower than a
ground-based solution of 0.1543 obtained by Young et al. (BAAS 25,
1137,1993). Values and error estimates for masses, densities and other
important solution parameters will be presented. This work was performed
with partial support from Space Telescope Science Institute.

28.08 Schmitt B.* Doute S. Quirico E. Benchkoura A. de Bergh C.
Owen T. C. Cruikshank D. P.
The State and Composition of the Surface of Pluto: Laboratory
Experiments and Numerical Modeling

The analysis of the near-infrared observations of Pluto (Owen et al.,
Science, 1993) made at UKIRT revealed that the positions of the methane
bands do not coincide with the positions of pure solid methane but
rather with methane diluted in solid nitrogen. Additional laboratory
work and a more detailled spectral analysis showed that the strong bands
above two microns are also not well matched in position by diluted CH4,
and that some differentiation may have occured (Schmitt et al., BAAS
1993). Using spectral reflectance modeling, we further analyzed the
positions, shapes and intensities of the various absorption bands in the
Pluto spectra in order to derive additional constraints on the state,
composition or temperature of the surface as well as possible
distributions of these.parameters. We also developed a numerical model
and performed a series of laboratory experiments simulating the
sublimation process of CH4:N2 ice mixtures on Pluto in order to
understand its spectroscopic and physico-chemical implications.

28.09 McKinnon W. B.* Brackett R. A.
Jetting and Ice Loss During Collisinal Formation of the
Pluto-Charon Binary

We have reexamined the sphere-on-sphere jetting model of Melosh and
Sonett (1986, Origin of the Moon) for colliding ice and rock-ice
spheres, using the icejetting parameters from McKinnon (1989, GRL 16).
We find that the maximum loss due to jetting, which occurs for the
symmetric collision of two equally sized spheres, is ~75% of an impactor
volume (or ~40% of the total). Included in the model are the explicit
behavior of the critical angle for jetting of water ice and a more
realistic scaling of the jet width to the shock penetration depth. The
relative loss of ice is remarkably similar to earlier results for
equivalent, high-speed rock-on-rock collisions, despite the critical
angle and jet width modirlcations; compensating changes are due to our
rederivation of the jetting model. In the context of forming
Pluto-Charon, the above upper limit applies to the collision of
differentiated proto-bodies. Forty percent volume loss, as ice, could
take Pluto-Charon from an ~55/45 rock/ice ratio (the best present
estimate of the kinetically inhibited, outer solar nebula value) to
~75/25 (greater than implied by present estimates of Pluto's radius).
For proto-body mass ratios other than unity, the amount lost to jetting
is smaller, but accounting for Pluto-Charon's high rock/ice ratio
through a single jetting event is plausible. Jetting losses are also
sensitive to velocity. For collision speeds greater than ~2.5 km/s,
appropriate to Pluto's inclined, eccentric orbit, the above results are
pertinent. For collisions at escape speed, ~1.3 km/s, most jetted ice
does not escape; only ~10% total volume is lost for the symmetric
collision of two equal spheres. Thus if jetting is the mechanism for
raising Pluto's rocklice ratio, a higher velocity collision is favored;
this is consistent with the angular momentum density of the binary and
the simulations of Levison and Stern (Icarus, submitted) for stabilizing
Pluto within the 3:2 resonance by a Charon-forming impact.

This research supported by NASA grant NAGWP432.

28.10-P Quirico E. Schmitt B.
Infrared Spectroscopy of Molecular Solids and of CO and CH4
Trapped in Nitrogen Matrix

The interpretation of near infrared observations at high spectral
resolution (< 10 cm^-l) of icy surfaces, such as the surfaces of Triton
and Pluto, positively needs a deep understanding of the physics of the
molecular solids suspected to be present. Infrared observations contain
a lot of information about both the physical and the chemical states of
the surface, such as composition, temperature, crystalline phase...

We developed several cryogenic cells giving us the possibility to grow
monocrystals with excellent crystalline quality (thicknesses from 25
micrometers up to 1 cm). We performed several systematic experimental
studies on pure molecular solids (N(sub)2,CH(sub)4, CO, ...) and on
molecules trapped in matrices (CO and CH(sub)4 in solid nitrogen) as a
function of temperature, concentration and thermal history. We analyzed
the origin of the bands, their dependency with different physical
parameters and observed the reciprocal effects of guest and host
molecules in the case of molecules trapped in the N2 matrix. First works
already allowed us to extract information about the surfaces of Triton
and Pluto: detection of CO and CO2 on Triton and N2 and CO on Pluto-
diluted state of CH4 (Cruikshank et al. 1993, Quirico and Schmitt 1994)

However, all information in the astronomical spectra is still far to be
extracted and we will present the results of further studies conducted
on pure N(sub)2, CH(sub)4, CO and CO(sub)2, and on these molecules
trapped in solid nitrogen.

28.11-P Hansen C. J. Paige D. A.
A Parameterized View of Pluto's Climate History

Seasonal nitrogen cycles are of interest on Pluto because this volatile,
which dominates the composition of Pluto's surface ices will also
control atmospheric pressure if Pluto's atmosphere, like Triton's, is in
vapor pressure equilibrium with its surface ices. Even at Pluto's cold
temperatures nitrogen will sublime and condense seasonally. What effect
will Pluto's high obliquity and eccentric orbit have on the location of
its frost deposits?

A thermal model developed for Triton has been modified and applied to
Pluto. The thermal model solves the frost energy balance equation for
nitrogen frost, balancing solar insolation, thermal emission, and heat
conduction between the substrate and the frost deposit, with frost
temperature changes buffered by the heat capacity of the frost deposit,
and the latent heat of condensation and sublimation. The alpha-betaphase
transition of solid nitrogen, which takes place at 35.61 K, is
incorporated in the heat balance equation at the transition temperature.
The primary input parameters to the model are the albedo and emissivity
of the frost, the albedo and thermal inertia of the substrate, and the
total nitrogen inventory. As on Triton, these parameters are poorly
constrained by observations and laboratory experiments. The model output
can be compared to recent Pluto observations. Maps of albedo features
from the Pluto-Charon mutual events are compared to the model's
predictions for polar cap boundaries. The atmospheric pressure derived
from the 1988 stellar occultation is compared to that calculated by the

Pluto's high obliquity is found to have a significant effect on the
distribution of frost on its surface. Conditions that would have led to
permanent polar caps on Triton were found to lead to permanent zonal
frost bands on Pluto. In some instances, frost sublimed from a seasonal
polar cap from the pole outward, resulting in a "polar bald spot". Dark
frost did not satisfy observables on Pluto, in contrast to our findings
for Triton. Bright frost came closer to matching observables, but was
not completely satisfactory.

Seasonal variations in atmospheric pressure were observed in the model
results, and the amplitude, and to a lesser extent the phase, of the
variations depended significantly on frost and substrate properties. In
most cases two peaks in atmospheric pressure were observed annually: a
greater one associated with the sublimation of the north polar cap just
as Pluto receded from perihelion, and a lesser one associated with the
sublimation of the south polar cap as Pluto approached perihelion.
Atmospheric pressure was thus found to be determined both by Pluto's
distance from the sun, and by the subsolar latitude.

Our model predicted frost-free substrate surface temperatures in the 50
to 60 K range, depending on the thermal inertia assigned to the
substrate, while frost temperatures typically fell between 30 to 40 K.
Temporal changes in frost coverage illustrated by our results and
changes in the viewing geometry of Pluto from the earth may help to
minimize the apparent differences in ground-based measurements of
Pluto's thermal emission.

SESSION 29 ....... Asteroids II
Friday, 10:30 - 12:00 Crystal Ballroom A
S. J. Ostro and F. Vilas, Moderators

29.01 Lovell A. J.* Schloerb F. P. Lagerkvist C.-I.
Lagerros J. S. V. Lumme K.
Observations of Ceres at 3.0 mm Wavelength

Measurements of the thermal radiation from asteroids at millimeter
wavelengths provide information about the thermo-physical properties of
their upper few centimeters. We have obtained observations of Ceres at a
wavelength of 3.0 millimeters using the Five College Radio Astronomy
Observatory 14-m millimeter-wave antenna. The observing strategy makes
use of a novel kind of millimeter-wave receiver, which enables
simultaneous observations of 15 separate sky locations. The use of this
imaging array receiver, known as QUARRY, substantially reduced many
systematic errors that plague typical continuum observations with single
radio antennas. Several measurements of Ceres were made during the
interval October, 1993-January, 1994 in order to sample the full range
of phase angle (5-20 degrees) that was available during this opposition.
Preliminary results of all observations are probably consistent, within
the likely statistical and calibration errors, with a disk brightness
temperature of 162 +- 19 K, normalized to a Ceres radius of 500 km.
However, there is marginal evidence for a decrease of about 20 K in the
disk brightness temperature over the range of phase angle observed.
These observations will be compared to thermal model calculations to
constrain the thermo-physical properties of the surface of Ceres.

29.02 Vilas F.*
Detecting Water of Hydration on Solar System Bodies

The 3.0 micrometers water of hydration absorption feature observed in
the IR photometry of many low-albedo and some medium-albedo asteroids
strongly correlates with the 0.7 micrometers Fe^2+-->Fe^3+ oxidized iron
absorption feature observed in narrowband spectrophotometry of these
asteroids. Using this relationship, an empirical algorithm for
predicting the presence of water of hydration in the surface material of
a Solar System body using photometry obtained through the ECAS v (0.550
micrometers), w (0.701 micrometers) and x (0.853 micrometers) filters
was developed and applied to the ECAS photometry of asteroids and
satellites. The percentage of objects in low-albedo, outer main-belt
asteroid classes that test positively for water of hydration increases
from P--> B-->C--> G class and correlates linearly with the increasing
mean albedos of those objects testing positively. The aqueous alteration
sequence in the Solar System thus ranges from P-class asteroids, which
represent the least-altered objects created at temperature_ attained at
the onset of aqueous alteration, to G-class asteroids, which represent
the upper range of the alteration sequence. The medium-albedo M-class
asteroids do not test positively in large number using this algorithm.
Spatial variation in the presence of water of hydration across the
surfaces of some asteroids is suggested in the variability of the water
of hydration test for observations of many asteroids on different
nights, although an effort to correlate variation with rotational period
is unsuccessful due to the small sample of eligible asteroids.
Aqueously-altered asteroids dominate the Solar System population between
heliocentric distances of 2.6 to 3.5 AU, bracketing the Solar System
region where the aqueous alteration mechanism operated most strongly. J
VI Himalia and Phoebe tested positively for water of hydration,
supporting the hypothesis that these may be captured C-class asteroids
from a post-accretional dispersion.

This work was supported by the NASA Planetary Astronomy program and
Origins of the Solar System Program.

29.03 Rivkin A. S.* Howell E. S. Britt D. T. Lebofsky L. A.
Determination of Near-IR Continua for M-asteroids

Based on their moderately high (0.15-0.25) albedos and red-sloped
featureless spectra, M-asteroids have been thought to be composed of
metals such as iron and nickel. However, observations done over the past
several years in the 3-micron region of the spectrum by Jones et al.
(1990 Icarus 88) and Rivkin et al. (1994 LPSC XXV) have shown at least
two and possibly three M-asteroids have surfaces containing hydrated
minerals. These observations imply not all M-class objects have
metal-rich surfaces.

Past work in the 3-micrometer region was largely concerned with
low-albedo asteroids, and observations were normalized such that
reflectances in the 3-micrometer spectral region less than at 2.2
micrometer indicated water-of-hydration absorptions (e.g. Jones et al.,
1990). This was justified because these low-albedo asteroids (of the C,
G and B classes) have rather flat continuum slopes throughout the region
of interest. Band depths could thus be simply calculated. The problem is
more complex for asteroids with large continuum slopes, however, and
because M-asteroids have a significant red slope compared to C-class
asteroids, it is necessary to calculate a continuum in order to decide
whether or not a feature exists. An absorption superimposed on a large
positive slope may still result in a positive slope from 2-3 micrometer,
so requiring a reflectance decrease from 2.2 to 2.95 micrometer may thus
be excessively conservative. Studies of the iron meteorite Mundrabilla
by Miyamoto (1987 Icarus 70) show continuum reflectivity is not a linear
function of wavelength but rather a function of (1/wavelength),
decreasing linearly with wavenumber from 2.5 micrometers to 25
micrometers. In the visible and near-IR, we also find reflectivity
decreasing linearly with wavenumber in Gaffey's spectra of iron
meteorites and enstatite chondrites (Gaffey, 1976 JGR 81).

In light of this, we performed a least-squares fit to spectrophotometric
data collected in 1991 and 1993 at the IRTF to obtain a continuum for
each asteroid, which we extrapolated to the 3-micrometer region. A
"continuum envelope" can be formed from the values for slope and zero
point one sigma below and above the best-fit value. The M-asteroids we
considered "anhydrous" are well fit by the extrapolated continuum, while
3-micrometer water-of-hydration features are confirmed on 92 Undina and
201 Penelope. The continua fit to M-asteroid data taken in 1991 and 1993
will be shown and discussed. The above procedure will work for
M-asteroids since their putative analogs (enstatite chondrites and
iron-nickel meteorites) have continua that decrease linearly with
wavenumber. This technique may also prove valuable in showing the
contributions of metals to a spectrum.

29.04 Xu S.* Binzel R. P. Burbine T. H. Bus S. J. Sunshine J.
Small Main-Belt Asteroid Spectroscopic Survey: Initial Results

The Small Main-Belt Asteroid Spectroscopic Survey (SMASS) has measured
the reflectance properties of 316 main-belt asteroids. The median
diameter for the sample is 20 km. The survey utilized a CCD spectrograph
attached to the 2.4-m and 1.3-m telescopes of the Michigan-Dartmouth-MlT
Observatory located at Kitt Peak, AZ. The spectra cover the wavelength
range of 4000 to 10000 Angstrom with a dispersion of 10 Angstrom/pixel.
Principal Component Analysis was used to classify the observed small
asteroid population and two new classes are justified, which we label as
J and O. The J-type asteroids appear linked with asteroid 4 Vesta and
may be asteroid analogs to diogenite (basaltic achondrite) meteorites.
The O-class asteroid, 3628 Boznemcova, has a spectrum resembling
ordinary chondrites. The majority of the small main-belt asteroids
belong to S and C classes, as is the case for larger asteroids. However,
small asteroids show more diversity in absorption features than larger
ones, which may indicate a greater variation of compositions in the
small asteroid population. This variation could arise if many small
asteroids are comprised of distinct compositional units from within once
larger parent bodies. Olivine-rich A-type asteroids are rare in our
observed size range, although a strong 1-micrometer absorption band is
present in the spectra of many small (D </- 20 km) S-type asteroids.
This strong absorption band suggests the possibility of a significant
olivine component, clinopyroxene component, or both. Extending the
wavelength coverage into the near-infrared will be necessary to resolve
the mineralogy of these small asteroids.

29.05 Clark B. E.* Hiroi T.
S-Type Asteroid Spectral Continua: Redness and Fe,Ni Metal

Combining evidence from recent spectral surveys and compositional
modeling, we show that our previous notions of inner asteroid belt
geology may have to be reconsidered. Spectroscopic interpretations have
long associated S-type asteroids, the most abundant class of asteroids
in the inner main belt, with various combinations of olivine,
orthopyroxene, and Fe,Ni metal (Pieters et al. 1976; Gaffey and McCord
1978). It is currently supposed that S-types are thus best represented
by the achondrite and anomalous stony-iron meteorites (McCord and
Gaffey, 1974; Gaffey, 1984; Bell ei al. 1985). With the following
reasoning however, we suggest that metallic components may not be
responsible for the redness of most S-type asteroid spectra, and that
their meteoritic spectral analogs are therefore yet open to question.

1) First, S-type asteroids have red spectral continua that increase from
0.7 to 1.5 micrometers across the mafic mineral absorption band at 1
micrometer. This slope has long been attributed to a Fe,Ni metal
component among the surface materials. 2) There is a distinct break in
continuum slope at 1.5 micrometers at which point the continuum levels
off from 1.5 to 2.5 microns. The cause of this break has been assumed to
be due to silicate components. 3) 90% of the S-type asteroids measured
for the 52-color (Bell et al. 1988) and SCAS infrared surveys (Clark et
al. 1994) show this break in continuum. Broadband JHK colors of S-types
(Veeder et al. 1978) also show this characteristic. 4) Metallic Fe,Ni
has a red spectral continuum that increases uniformly and almost
linearly with wavelength from 0.5 to 2.5 micrometers (Gaffey, 1976,
1986; Britt and Pieters, 1988). 5) S-type asteroids have been
compositionally modeled with achondrite meteorites and the Mundrabilla
Fe,Ni meteorite, but with only 40% success (Hiroi, et al. 1993). 6)
S-types have also been modeled with pure mineral endmembers. However,
the success rate is equally ambiguous (Clark, 1993). 7) The break in IR
continuum slope from 1.5 to 2.5 micrometers is the main cause of
distress in modeling S-asteroids with mixtures of silicate and Fe, Ni
metal endmembers (Clark, 1994; Hiroi et al. 1993). i.e. The silica
component spectral turnover at 1.5 micrometers is insufficient to mask
the spectral redness from 1.5 to 2.5 micrometers of metal components
included to account, for the redness from 0.7 to 1.5 micrometers. It may
be that S-types have other important phases that could account for the
continuum turnover. For example, spinel and cpx have been suggested, but
not yet investigated (Hiroi, 1994).

It becomes increasingly clear that mineralogic interpretations of S-type
spectral continua are not so trivial as supposed. The bad news is that
we still don't know why S-types are red. The good news is that spectral
continua may be compositional information after all. Among the
implications of this suggestion is the radical idea that we do not
necessarily know what kind of rock comprises most of the geology of the
inner asteroid belt.

29.06 Bell J. F.*
Space Weathering on S-Class Asteroids: Myth vs. Reality

MYTH: In the 1960s meteoriticists expected on the basis of meteorite
fall statistics, cooling rates, and collisional evolution models that
the asteroid belt would be dominated by ordinary chondrite material at
all sizes. When the first spectroscopic surveys in the early 1970s did
not find any OC-like objects among the larger asteroids, the failure of
the original prediction was rationalized by invoking a "space
weathering" process that camouflaged the hypothetical OC asteroids with
spectrally altered regoliths that exhibit S- or C-type spectra. None of
the original reasons for expecting large OC asteroids are still valid.
Modern meteorite cooling models suggest that the OC parent bodies were
never more than ~SOkm in diameter. The steep crater abundance curve on
Gaspra shows that the uniform power law of index -3.5 predicted by
classical collisional evolution models for asteroids does not exist. The
true asteroid size distribution is now believed to follow the complex
post-mare lunar cratering function. It now appears possible that
chondrites and irons have very different size distributions, breaking
the assumed link between meteorite fall statistics and asteroids.
Furthermore, a genuine OC (Class Q) asteroid has been found in the
asteroid belt at the solar distance and size predicted by a model
assuming no space weathering (Bell et al., Asteroids 11, p. 942). The
Galileo images of Ida show an extremely uniform surface, with no fresh
crater bottoms showing OC-like spectra. Meteoriticists now recognize a
much larger number of differentiated meteorite classes which are good
alternate analogs for S asterords. All these discoveries are strong
arguments against the "classical" space weathering model.

REALITY: However, there is now strong evidence for a very different sort
of space weathering on S-type asteroids, one which mostly affects
silicate band depth rather than continuum slope and curvature. It has
long been observed that small Earth-crossing asteroids tend to have deep
bands relative to large main-belt objects, but it was not clear if this
was related to size or orbital history. Recently it was shown that
main-belt S asteroids below ~100km diameter show a smooth function of
decreasing band depth with increasing size (Gaffey et al., ICARUS 106,
592). This effect is found in all S-types, independent of Ol/Pyx ratio
or other mineralogical parameters, and it appears to "saturate" near
100km diameter with all S asteroids above this size having roughly
similar band depths. These features suggest the existence of a change in
regolith properties on all S-type asteroids with surface gravity and/or
age. The differences between Gaspra and Ida in crater density, spectral
uniformity, ad thermal inertia are consistent with such a change. This
effect, rather than intrinsic mineralogical differences. probably
accounts for most of the spectral differences between Ida and its
satellite seen by Galileo.

29.07 Lazzarin M.* Barbieri C. Di Martino M. Barucci M. A.
Fulchignoni M.
Visible Spectroscopy of Dark, Primitive Asteroids

Asteroids and comets have a close relationship with the planetesimals
which formed from the solar nebula 4.57 billion years ago. Since then,
comets have been submitted to very low levels of evolutionary pressure
and the differentiation of some asteroids has been spread over at most
the first few million years. As such, small bodies have preserved
material which witnessed the condensation and the early phases of the
formation of the solar system. In particular, speaking about asteroids,
it is now evident that the darkest objects are also the most primitive.
Laboratory spectra suggest that the sources of the low-albedo matter may
be a mixture of organic compounds together with acqueous alteration
products of igneous minerals. Therefore, these asteroids may be the most
closely related to comets.

In this context we started in 1993 an observational program of dark
asteroids, precisely those of taxonomic class C, D and P to investigate
spectroscopically these objects in order to study their surface
composition and make a comparison among the different taxonomical
classes, to reveal possible weak features of cometary origin to learn
about the possible link between comets and asteroids. The data, 20 low
resolution spectra in the visible range, have been obtained during
different observational runs carried out at the Observatories of La
Silla-Chile and of Asiago-Italy. The results of the survey will be
presented and discussed.

29.08 Howell E. S.* Rivkin A. S. Marcialis R. L. Nolan M. C.
Grier J. A. Bus S. J. Slivan S. M.
Spectral Observations of Near-Earth Asteroids

Over the past year, we have studied the composition of near-Earth
asteroids. Using both visible and near-infrared photometry and
spectroscopy, we have obtained physical data for 15 near-Earth asteroids
to assign taxonomic classifications and determine surface mineralogy. If
the apparent discrepancies between asteroid spectra and meteorite
spectra are due to regolith processes that affect the spectral slope,
albedo, and band depth, these effects may be less prevalent on smaller
near-Earth asteroids than on main-belt objects. Near-Earth asteroids are
the more immediate parents of meteorites, at least in a statistical
sense, and may be more similar to meteorites in spectral appearance.

Two near-Earth asteroids 4953 1990 MU and 4954 Eric, both determined to
be S class, had particularly favorable apparitions in 1994. We
coordinated simultaneous visible and near-infrared observations for 4954
Eric. Spectra in both wavelength regions were also obtained for 4953
1990 MU, although separated by several days. These objects both show a
pyroxene absorption band centered near 2 micrometers. The near-infrared
spectra were obtained using the F. L. Whipple Qbservatory Mulitiple
Mirror Telescope (MMT) and the FSPEC near-infrared spectrometer. Most of
the CCD spectra were taken with the 1.54-m Catalina Station telescope.
However, the simultaneous visible observations of 4954 Eric were
obtained at the 2.4-m Michagan-Dartmouth-MIT telescope. We combined
visible and near-infrared photometry taken over a time period of less
than 2 hours from the McGraw-Hill telescope and MMT, respectively, to
scale the CCD and near-infrared spectroscopy into a composite
reflectance spectrum of 4954 Eric. Analysis of these spectra and the
compositional implications will be presented.

29.09 Rabinowitz D.* Howell E. Hicks M.
Unusual Colors for Small Earth Approachers

In March and April of 1994, seven new Earth approachers with sizes
ranging from 5 to 75m were discovered at the Spacewatch Telescope of the
University of Arizona during a search optimized to detect small Earth
approachers (SEAs). Two of these objects (1994 ESl and 1994 GV) were
observed spectroscopically (0.5 - 0.9 ,micrometers), and one (1994 EU)
was observed through Harris B and V, and Cousins R filters (0.44, 0.55,
and 0.67, micrometers). These new observations, and previous color
measurements for five other SEAs (1991 VA, 1991 VG, 1992 DU, 1992 JD,
and 1993 KA), show that the reflectance spectra of SEAs differ markedly
from the reflectance spectra of main-belt asteroids and larger Earth
approachers. Some of the SEA spectra are unusually reddened at
wavelengths longward of 0.55 um, and a few show sharp absorptions
shortward of 0.55 um. Unfortunately, most of these spectra are not
sampled over a wide enough wavelength range to compare conclusively with
meteorite samples. However, it is clear that SEAs originate from diverse
sources, and that some of these sources have not been observed
spectroscopically among the main-belt asteroids or larger Earth
approachers. Possible sources, consistent with the sources determined by
analysis of the SEA orbits [1] are cometary debris, and impact ejecta
from the Moon, Mars, Earth, or an undiscovered population of Earth

[1] Rabinowitz, D.L., 1994, Icarus, in press.

This work was supported by NASA, and would not have been possible
without the generous support of the entire Spacewatch team.

29.10-T Vasilyev S. V.
System Analysis of Telescope and Laboratory Polarimetric Data
for Rough Surfaces

The attempt to describe the polarization phase curves of atmosphereless
bodies and laboratory data is made by principal component method. The
data include the polarization phase curves of more than 50 asteroids,
some comets and various laboratory data including the meteorites, lunar
and artificial samples. Linear independent parameters completely
describing all the differences among the initial curves have been found.
Two and three-dimensional plots of these parameters (eigenvalues) show
the asteroid and comet grouping in accordance with their belonging to
composition types. Such diagrams can be used for independent
determination of asteroid types. On the other hand the laboratory
samples revealed the sensitivity of the method to physical changes in
examined surface. It was found that some parameters have a significant
correlation with physical parameters. The small number of obtained
parameters and their linear independence has allowed to provide more
correct comparison of telescope and laboratory data. The results of
comparative analysis are discussed.

29.11-P Muinonen K.
Light Scattering by Stochastically Rough Particles

The geometric optics approximation for light scattering by
stochastically rough particles has been studied by Peltoniemi et al.
(Appl. Opt. 28, 4088, 1989); the present report is a natural
continuation of their work. I describe three-dimensional stochastic
particle shapes by multivariate lognormal statistics and compute full 4
x 4 scattering phase matrices, thus improving significantly on the
earlier approach. The stochastic geometry can be generated by using
spherical harmonics expansions having random variable coefficients that
satisfy certaln statistical conditions. A new autocorrelation function
has been devised for the particle radius, allowing an analytical
derivation of the spherical harmonics coefficients. Once the particle
geometry has been generated, scattering in the geometric optics
approximation can be computed via rather general ray tracing algorithms
for an arbitrary particle shape r = r(theta,phi) theta and phi being
spherical coordinates). I present scattering phase matrices for a large
selection of particle refractive indices and statistical shape
param-eters; in particular, I compute asymmetry parameters both
including and excluding the forward diffraction component. The results
help us understand light scattering by solar system dust particles, and
thereby constrain the physical properties of, for example, asteroid
regoliths and cometary comae.

29.12-P Hudson R. S. Ostro S. J. Harris A. W.
Asteroid 4769 Castalia: Interpretation of Optical Lightcurves
Using a Radar-Derived Shape Model

R. S. Hudson (Washington State U.), S. J. Os?o, A. W. HalTis

A 167-pararneter, 3-D shape model of Castalia [1]. obtained from
inversion of delay Doppler images, constrained the object's subradar
latitude on the date (Aug. 22,1989) of the observations by defining an
annular sky region that contains Castalia's pole. We have used the shape
model and norninal Hapke parameters for S asteroids[2] to model Castalia
lightcurves obtained at Table Mountain Observatory on Aug. 23-25, 1989,
at solar phase angles ranging from 58 degrees to 90 degrees. This
investigation defines the projected pole's azirnuthal orientation around
the radar lille of sight and yields a least-squares estimate of the
orientation of Castalia's north or south pole (within 10 degrees of
lambda = 250 degrees, beta = -40 degrees). There is insufficient
geometric leverage to distinguish the rotation sense or to estimate
values of Hapke parameters. The model lightcurves match the TMO data
surprisingly well, given the modest number of shape parameters in the
model and the fact that part of the asteroid was not seen by the radar.

Visual comparison of the lightcurves with the frame-by-frame geocentric
appearance of the sun-illuminated shape model reveals the detailed
relationship between the asteroid's visible, illuminated, projected
surface area and its disc-integrated brightness. For example, one can
see why one lightcurve minimum is deeper and sharper than the other and
why one maximum is weaker and flatter than the other -- small
asymmetries in Castalia's overall shape and in the character of
concavities at various scales play roles in determining the form of a
lightcurve under any given viewing/illumination geometry. The
availability of a detailed shape model clearly enhances the
interpretability of lightcurves, especially for very small (and hence
potentially very irregular) obJects observed at large phase angles.

In their inversion of the radar images, Hudson and Ostro presented
models corresponding to lower and upper bounds on Castalia's
bifurcation. The lightcurve data clearly favor the more bifurcated
model. Our experience suggests that optical lightcurves may enhance the
accuracy of future radar reconstructions of small asteroid shapes, to a
degree that would depend on the geometrical circumstances encountered in
the radar and optical observations as well as on the asteroid's shape
and spin state. For example optical determination of a target' s the
spin vector would shrink the parameter space for the radar inversion. ln
some cases, lightcurves might provide extra leverage in splitting the
N/S ambiguity in single delay-Doppler images.

[1] R. S. Hudson and S. J. Ostro 1994, Science 263, 940-943.
[2] P. Helfenstein and J. Veverka 1984. In Asteroids 11 (R. P. Binzel,
T. Gehrels, and M. S. Matthews, eds.). pp. 557-593, Univ. of Arizona,

29.13-P Verbiscer A. Veverka J.
The Effects of Asteroid Shape and Surface Roughness on Spectral

We use Galileo imaging data of Gaspra and Ida as well as complementary
groundbased observations obtained in support of the Galileo encounters
to assess the importance of surface roughness and body shape effects in
modifying S-asteroid reflectance spectra in the 0.4-1.2 micron region.
Over a decade ago Gradie, Veverka, and Buratti (1980, Proc. Lunar Sci.
Conf. 11th, 799-815) demonstrated that spectral slopes and the depths
and shapes of absorption features can depend on photometric geometry and
concluded that at some level, an asteroid's detailed spectrum will
depend on the shape, topography, and surface roughness of the asteroid.
With the availability of the Hapke photometric theory (Hapke, 1986,
Icarus 67, 264-280) and detailed shape models of Gaspra (Thomas et al.,
1994, Icarus 107, 23-36) and Ida, we investigate the types and amounts
of spectral modification that can be produced by roughness and shape
effects at different phase angles and under different rotational
aspects. Our results assess the degree to which small spectral
variations among similar S-asteroids, and spectral variations with
rotational aspect of the same asteroid, can be interpreted safely as
indicating variations in surface texture or composition.

29.14-P DeSanctis M. C. Caruso A.
Asteroid Complex Rotation

Recently, the observations the Apollo asteroid 4179 Toutatis have
pointed out the problem of the complex rotational state of asteroids.
Harris (Icarus, 1994) has re-examined the damping timescale of the
rotational wobble for asteroid and he has identified several objects
which might be expected to be in a complex rotational state (31021981
QA, 4179 Toutatis, 1220 Crocus, 288 Glauke, 1689 Floris-Jan, 3288
Seleucos and 887 Alinda). Among this group, 4179 Toutatis is most
notable: observations suggest that its rotational motion is more complex
than the simple principal axis rotation. We have tested the possibility
for Toutatis of multiple periods, as described in Spencer et al., (1994,
Icarus in press). To analyze the possibility of complex rotation with
multiple periods, a program based on the Phase Dispersion Minimization
(PDM) technique (Stellingwerf, Ap. J., 224, 953, 1978) usually used for
variable stars, has been developed. Applying this algorithm to the
Toutatis data, double periods appear to be possible. Toutatis is the
first asteroid to show strong evidence of complex rotation.

The asteroids Floris-Jan and Seleucos, are the other best candidates for
demonstrated wobble motions. Now, we are analyzing the observations of
Floris-Jan in terms of complex rotation taking into account that the
wobble period for this asteroid should be several times longer than the
spin period.

29.15-P Keesey M. S. W. Chodas P. W. Yeomans D. K. Wimberly R. N.
Predicted Earth Approaching Asteroids and Comets During the
Years 1995-2005

A table of asteroids and comets that will approach Earth to within 0.5
AU during the 1995 to 2005 time interval is presented. Each entry gives
the close approach date, minimum distance, relative velocity, apparent
magnitude, and declination. The declinations 10 days prior to and
following close approach are also included. These results were obtained
by numerically integrating the orbits of over 280 asteroids and 55
comets whose perihelion distances are less than 1.5 AU. These objects
were selected from our database of about 6000 numbered asteroids, 4000
unnumbered asteroids, and 150 short period comets. In addition to
optical data, radar was employed in determining the orbits of many of
the Earth approaching objects.

During the 11 year interval there will be over 350 occurrences in which
an asteroid or comet will pass within 0.5 AU of Earth and over 120 in
which an object passes within 100 lunar distances (0.257 AU). Asteroid
4179 Toutatis will make the closest approach to Earth, in September
2004, with a minimum distance of about 4 lunar distances.

In addition to the close approach table, a table of orbital elements of
the Earth approaching objects is provided. Ephemerides for any of the
objects will also be provided upon request.

29.16-P Whipple A. L.
Chaos Among the Planet Crossing Asteroids

The maximum Lyapunov characteristic exponent (LCE), and the
corresponding Lyapunov time, is a measure of the exponential divergence
of neighboring orbits in phase space. The orbits of asteroids all have
Lyapunov times that are short relative to the orbits of the planets. The
planet crossing asteroids have extraordinarily short Lyapunov times,
virtually all are less than 5000 years.

The results of numerical integrations of the orbits of all of the
numbered asteroids whose current perihelion distance is less than 1.6 AU
are presented. These integrations included the perturbations of Mercury
through Pluto. The state transition matrix was integrated along with the
equations of motion and the Lyapunov time was calculated from the state
transition matrix. The durations of the integrations were determined by
the growth of the state transition matrix. The integrations were
terminated when the estimate of the maximum LCE converged or at 150,000
years, whichever occurred first.

The Lyapunov time is shown to depend very strongly on perihelion
distance. The Lyapunov time and the other orbital elements are generally
independent, with the exception of a weak correlation between
inclination and Lyapunov time. These results support that conclusion
that the predominant cause of the extreme chaos among the planet
crossing asteroids is their close encounters with the terrestrial
planets. Mean motion and secular resonances play a much weaker role in
the evolution of these objects once they are on planet crossing orbits.
The magnitude of the chaos found in these asteroids severely limits our
ability to predict their motions for even centuries.

29.17-P Slivan S. M. Binzel R. P.
A Spin Vector Survey for the Koronis Family

A long-term survey of lightcurves for members of the Koronis family is
being completed in preparation for an analysis of their distributions of
shapes and spin vectors. Through these distributions, it is our goal to
evaluate and constrain models for the age and collisional evolution of
this family, which contains the Galileo flyby target asteroid 243 Ida
and its satellite. The observational database includes published and
unpublished rotational lightcurves obtained through 1994. The most
recent data (1992-1994) were obtained using CCD imaging systems during
111 nights at the MIT Wallace Astrophysical Observatory, 5 nights at the
University of Texas McDonald Observatory, and 2 nights at the
Michigan-Dartmouth-MlT Observatory. Our shape and pole solution set will
consist of 12 Koronis family members. For each of these, we have
lightcurve measurements over widely spaced ecliptic longitudes from
typically three or more oppositions. These new lightcurves have already
yielded refined rotation periods for 7 of these asteroids.

29.18-P Terrell D. Durda D. D.
Inversion of Asteroid Lightcurves by Differential Corrections

We present a versatile and generalized method for inverting asteroid
lightcurves which extracts a maximum of physically relevant information,
and which may be conveniently used by researchers in the field of
asteroid photometry. Our approach is to compare the modeled lightcurves
of an asteroid with a specified shape and spin state to a number of
observed lightcurves. The shape parameters and spin state of the model
asteroid are adjusted by the method of differential corrections until an
impersonal fit to all lightcurve data is achieved. The computer code
utilizes a modularized design, allowing the user to apply the method
with any shape model and scattering law. The model also allows for later
incorporation of occultation data into the simultaneous solutions,
placing a very powerfill constraint on the shape of the asteroid being
modeled. This ability to incorporate different types of observations
proves to be extremely important to the quality of the solutions,
because different types of observations each have their own strengths
and weaknesses. A simultaneous solution of multiple data types makes
possible the recovery of parameters that are impossible to recover or
only weakly recovered by a single type of observation. Our goal is to
make the program as general as presently possible and extendible to
future improvements in accuracy and observations.

29.19-P Larson H. P. Montani J. Lebofsky L. A. Rivkin A. Howell E.
Moran C.
Volatiles in Asteroidal Surface Minerals

We are analyzing the reflectance spectrum of asteroid 1 Ceres between
2.4-4.0 micrometers to clarify the role of trace volatiles in asteroidal
surface minerals. This spectral region includes the prominent signature
of H20 in hydrated minerals and more subtle features due to H20 ice, OH
in hydroxylated mincrals, CH in organic material, and NH(sub)4+ in
ammoniated minerals. The hydrate band has been detected in more than 3()
asteroids, including Ceres, but evidence for some of the other materials
has been reported only for Ceres because of the difficulty in getting
spectra with high resolution and high SNR for low-albedo asteroids. A
weak, partially resolved feature in Ceres' spectrum at 3.07 micrometers
was attributed to H2O iGe by Lebofsky et al. (1981) and to ammoniated
phyllosilicates by King et al. (1992). Either assignment has important
implications for the processing of volatiles in planetesimals, but
neither interpretation'is uniquely constrained by available data. We
have therefore assembled the best composite near-IR spectrum of Ceres
from available observations to reexamine this assignment. We have also
recorded laboratory comparison spectra of candidate materials at high
resolution over the entire near-IR spectral region to reveal all
possible diagnostic features, whether accessible or not in the
astronomical observations now available,-that could relate to the origin
and evolution of volatiles in asteroids. In this paper we compare the
composite near-IR spectrum of Ceres to these candidate materials, both
singly and mixed in various proportions, to demonstrate that its surface
cannot be modelled by hydrated minerals alone. We also reexamine the
assignments of H20 ice and ammoniated minerals to the feature at 3.07
micrometers, but we cannot convincingly choose between them. Other types
of observations and recent theoretical work that influence the
interpretation of Ceres' near-IR spectral reflectance will be reviewed.

29.20-P Chamberlin A. B. McFadden L. A. Schulz R. Schleicher D. G.
4015 Wilson-Harrington, 2201 Oljato, and 3200 Phaethon: Search
for CN(0-0) Band Emission

Our search for evidence of remnant cometary outgassing was uniquely
motivated for each of these unusual asteroids. Asteroid 4015 1979 VA was
discovered to be the same object as comet P/Wilson-Harrington 1949 III
and subsequently named 4015 Wilson-Harrington. 2201 Oljato is a
planet-crossing asteroid which may have evolved from either the main
asteroid belt or the reservoir of short-period comets into its current,
dynamically unstable, planet crossing orbit. There is observational and
dynamical evidence that this asteroid might be an inactive cometary

Until recently, meteor streams were exclusively associated with comets.
3200 Phaethon is the first asteroid to be associated with a meteor
stream. Because there are plausible mechanisms for asteroidal generation
as well as cometary generation of meteor streams, the pedigree of
Phaethon is uncertain. Results from two prior observations conflict: one
matches a typical S-type spectrum while the other shows unusually blue
spectral color (neither asteroidal nor cometary in nature). With
uncertainty in both origin and spectral characteristics, we decided to
observe 3200 Phaethon again.

Spectral photometric observations were obtained to search for the
CN(0-0) band fluorescent emission around 3880 Angstrom because it is the
most easily observed indicator of cometary outgassing. We used the OSU
long slit spectrograph on their 1.8m Perkins telescope at Lowell
Observatory. 4015 Wilson-Harrington and 2201 Oljato were observed in
October 1992 and 3200 Phaethon was observed in October 1993.

We present the 3-sigma upper limits on the production rate of CN
(Q(sub)CN) and, incorporating the known diameters, the percentage of
active surface required to generate this production rate. Upper limits
for 3200 Phaethon will be presented. The upper limit on Q(sub)CN for
2201 Oljato has been reduced from previous measurements by a factor of
60. Log QCN is <22.3 and <22.6 for 4015 Wilson-Harrington and 2201
Oljato respectively. These limits are lower by an order of magnitude
than QCN observed in low-activity comets.

29.21-P Britt D. L. Rivkin A. S. Howell E. S. Lebofsky L. A.
Observations of "Dry" C Class Asteroids

Observations have shown that about 40% of C class asteroids are
anhydrous, suggesting major differences in mineralogy within the C
spectral class. The reflectance spectra of these asteroids lack the
strong signature of the spectral absorptions of water and structural OH
near 3.0 , micrometers. The suggested meteorite analog for the "wet" C
asteroids are the hydrated CI and CM carbonaceous chondrites, but
analogs for the "dry" C asteroids are harder to find. The anhydrous CO
and CV carbonaceous chondrites are thought to analogs for the K
asteroids. Bell (1992) has suggested that the anhydrous C asteroids may
have silicate mineral assemblages closer to a dehydrated CM carbonaceous
chondrite and named these hypothetical meteorites CM3. Britt (1991) has
suggested that regolith processes can lower the albedo and mask the
absorption features of ordinary chondrite material and that some
ordinary chondrite parent bodies may have the spectral characteristics
of C asteroids.

We are reporting on the results of observations of C class asteroids at
NASA's Infrared Telescopic Facility on Mauna Kea and the Smithsonian
Astrophysical Observatory's Multiple Mirror Telescope on Mt. Hopkins.
These observations are part of a program to study the spectral variation
within the inner asteroid belt C class asteroids. This includes
observations of their near IR spectral continuum slope, any silicate
spectral features in the near IR, and, using the IRTF, their hydration
state as shown by the 3.0 micrometers absorption.

29.22-P Plath J. Sykes M.
Thermal Modeling of Triaxial Ellipsoids

A simple expression for the disk integrated thermal emission from a
triaxial ellipsoid with arbitrary orientation and observational phase
angle has been derived. We have also developed a numerical model of the
corresponding disk integrated brightness at visual wavelengths assuming
surface scattering properties described by Hapke's equations and
"asteroidal" Hapke parameter values. Together these are used to
stimulate asteroid observations, which are then analyzed using the
asteroid Standard Thermal Model, which assumes asteroids to be
spherical. The variation of the inferred diameter and albedo as a
function of rotational and observational phase are explored and compared
with "true" values. Though the STM has been calibrated against
occultation measurements of the large asteroids Ceres and Pallas, this
analysis provides insight into the intrinisic reliability of the STM
when applied to smaller irregular objects.

29.23-P Storrs A. D. Zellner B. Wells E. N. Weiss B. P. Tholen D.
Sichitiu R. Kowal C. T. Burleson W.
Imaging Observations of Asteroids from the Hubble Space
Telescope (HST)

Images of twelve asteroids have been obtained with the Hubble Space
Telescope. In some cases, the asteroid is resolved, but no companions
(satellites) have been detected. We present the "raw" images along with
versions restored using the maximum entropy method.

Asteroid sizes and brightness are compared to determine visible
reflectivity (albedo). These values are comparable to those obtained
through infrared radiometry. Direct determination of asteroidal size and
albedo has relied primarily on occultation observations and spacecraft
flybys, until the HST observations were made. The observations presented
here directly observe sizes and albedos of the illuminated portion of a
number of bodies. In general, the asteroids are only barely resolved, so
no meaningful constraints on shape and albedo variegation across the
disk can be found. In addition, the sizes determined by image
restoration appear to be consistently under that given by Tedesco et al.
(1989, Asteroids II, 1090-1138). This effect may be an artifact of the
image restoration.

Image restoration can be used to improve spatial resolution. In general,
the asteroidal data has been restored to twice the spatial resolution at
which it was obtained. This procedure also improves the chances of
retaining faint nearby images in the restoration process.

This work is based on observations with the NASA/ESA Hubble Space
Telescope obtained at the Space Telescope Science Institute, which is
operated by the Association of Universities for Research in AStronomy,
Inc. under NASA contract NAS5-26555. Some of this work was supported by
STScI grant GO-4521.01-92A.

SESSION 30 ....... Triton
Friday, 10:30 - 12:00 Crystal Ballroom B
W. B. McKinnon and L. M. Trafton, Moderators

30.01 Grundy W. M.* Fink U.
Band-Broadening Effects in Bright Planetary Surfaces

The initial phase of planetary spectroscopy involved measuring albedo
spectra and identifying the species responsible for producing the
absorption bands detected. This has now been done to the few-percent
level in the visible and near IR for all large, high-albedo bodies of
the Solar System, with absorptions due to H(sub)2O, SO(sub)2, CH(sub)4,
N(sub)2, CO, and CO(sub)2 ices being identified.

The next phase is to interpret more subtle spectral details in terms of
physical and chemical surface properties. Laboratory measurements have
shown that absorption band shapes and strengths can depend on
temperature (eg. Grundy et al. 1993, Tryka et al. 1993) and chemistry
(Schmitt et al. 1993). We have also found many mechanical configurations
in which absorption at weakly absorbing wavelengths is enhanced relative
to more strongly absorbing spectral regions. These effects change the
shapes of absorption bands, and can mislead by mimicking thermal and
chemical spectral changes. Three examples of band-broadening
configurations are described below. We expect that examples of all three
could be found on any large, solid-surfaced body in the Solar System.

A simple band-broadening source is the coexistence of geographical
terrains with different absorption properties. Imagine the spectrum of
an unresolved planet if parts of its surface show a strong absorption
band while other parts absorb weakly. The albedo contributions of
different terrains combine linearly with area, and, if the contrast is
large, the strongly absorbing geographic regions control the albedo at
weakly absorbing wavelengths while weakly absorbing regions control the
albedo at strongly absorbing wavelengths.

Grain size effects can lead to significant band broadening as well. For
power-law grain size distributions, such as dN = C r^-q dr, from which
the cumulative grain size law N(r) = C'r^1-q can be derived, there is a
value of the exponent, q ~ 3.5, at which maximum band-broadening occurs.
This value is between q = 4, where the total volume in each logarithmic
grain-size bin is the same and where the smallest grains begin to
dominate the spectral reflectance, and q = 3, where the total surface
area in each size bin is the same and larger grains begin to control the

Vertical inhomogeneity of surface properties can also lead to band
broadening. Though not easily modeled using the Hapke theory (Hapke
1981, 1984, 1986), Monte-Carlo models can reveal the spectral properties
of arbitrary three-dimensional configurations. A layer of small grains
above larger grains can produce band broadening, as can the reverse
situation, if the layers are thin enough to allow significant numbers of
photons to sample the subsurface layer.

This work was supported by NASA grants NAGW 1549 and NGT 50611.

30.02 Brown R. H.* Anicich V. G. Tryka K. A.
Heavy Isotopes of Carbon and Nitrogen on Triton and Pluto

The abundance of heavy isotopes of carbon and nitrogen in ices on the
surfaces of Triton and Pluto can provide important constraints on many
aspects of the evolution of these bodies. In particular, isotopic
fractionation occurring during the outgassing of volatiles from the
interior, as well as fractionation that occurs during atmospheric
escape, may provide a "window" into past evolution of the atmospheres,
surfaces and interiors of Triton and Pluto. In addition, isotope
fractionation may occur as seasonal volatile transport acts to
preferentially evaporate and condense lighter isotopic molecules,
increasing the concentration of heavier isotopic molecules in areas
experiencing a net volatile loss, while reducing the concentration of
heavier isotopic molecules in areas experiencing a net volatile gain.
Modern infrared instrumentation has now reached a level of sensitivity
that allows the detection of the signatures of the heavy isotopes of
carbon, nitrogen and hydrogen, among others, in groundbased reflectance
spectra. Examples include the possible detection of ^180 in SO(sub)2 ice
on Io (Howell et al., Icarus 78, 27-37) and the tentative detection of
^13C in CO ice on Triton (Cruikshank et al., Science 261, 742-745). As a
result, we have initiated a program to measure the near-infrared optical
constants of samples of ices of N(sub)2, CO and CO(sub)2, which contain
significant concentrations of the heavy isotopes ^15N and ^13C. We are
using these measurements to analyze existing spectra of Triton and Pluto
in an attempt to constrain the abundance ratios ^13C/^12C and ^15N/^14N
on Triton. Results of this work, and possible implications of the
isotopic abundance ratios will be discussed.

30.03 Khare B. N.* Sagan C. Heinrich M. Thompson W. R.
Arakawa E. T. Tuminello P. S. Clark M.
Optical Constants of Triton Tholin: Preliminary Results

Titan tholins produced at sub-mb pressures by charged particle
irradiation of simulated Titan atmospheres ([CH4]~0.1) exhibit optical
constants consistent within probable errors with the Titan haze [Sagan
et al. (1992) Acct. Chem. Res., 29, 286]. No other proposed material
comes close. Triton also possesses a N2/CH4 atmosphere, with P(sub)s ~16
microbars and [CH4]~10^-5 [Broadfoot et al. (1989) Science, 246, 1459.]
Triton and Titan tholins represent different cases with remarkably
different organic chemistry in the continuum of N2/CH4 irradiation
products [McDonald et al. (1994) Icarus, 108, 137] so Triton tholin is
of interest in its own right. Also, some e^--synthesized Triton
atmosphere tholin is expected, and may model the more abundant products
from irradiation of the corresponding surface ices [Thompson and Sagan
(1990) 250, 415]. Laboratory synthesis of large amounts of Triton tholin
is difficult because of the low [CH4}. We have successfully produced
sufficient quantities with a plasma discharge through 0.1% CH4 in N2 at
0.86 mbar and by achieving a leak-tight system preventing destruction of
slow forming Triton tholins by O atoms in the plasma discharge. Using
our previously tested thermal evaporation techniques we produce optical
quality films of varying thickness from mg quantites of Triton tholin.
The films had a slight brownish tint. The optical constants N(lambda)
and K(lambda) were determined spectroscopically. Preliminary analysis
indicates that K(lambda) decreases by an order of magnitude from ~10^-1
at 0.2 micrometers to ~10^-2 at 0.4 micrometers; from 0.4 micrometers to
1.0 micrometers, there is a slight decrease; k(lambda)is almost flat
from 1.0 micrometers to 2.5 micrometers. The complete analysis of
n(lambda) and k(lambda) from 0.06 micrometers to 40 micrometers will be

30.04 Allen M.* Dissly R. Anicich V.
Photolysis of Ethylene Ice on Triton's Surface

We have performed a number of UV photolysis experiments on C2H4 ice to
simulate the irradiation of this species on the surface of Triton.
Photochemical models of Triton's atmosphere predict C2H4 as a primary
product of methane photodissociation, with a deposition rate sufficient
for it to be readily observable as a surface condensate in <10^6 years
(Lyons, et al., 1993, submitted to Icarus), yet it has not been
observed. Our results show that C2H4 ice is decomposed by radiation at
wavelengths <1849 angstrom, with C2H2 (acetylene) as the primary
product. Dilution of the C2H4 in an N2 matrix does not appear to affect
the photochemical yield of C2H2, suggesting that the reaction pathway
for C2H4 ---> C2H2 is due to unimolecular dissociation. Quantum yields
for both the destruction of C2H4 and the formation of C2H2 will be
discussed, as functions of both irradiation wavelength and dilution in
N2 ice, which determine whether the ambient UV flux reaching Triton's
surface is adequate to explain the non-detection of ethylene ice.

This work represents research carried out at the Jet Propulsion
Laboratory, California Institute of Technology, under contract to the
National Aeronautics and Space Administration.

30.05 Pisano D. J.* Stansberry J. A. Spencer J. R.
The Emissivity of N2 Ice: Hapke Theory

Thermal balance studies based on Voyager 2 imaging data (Stansberry et
al., JGR, 1990; Hillier et al., Science, 1990; Stansberry et al.,
Icarus, 1992) consistently predict blackbody temperatures for Triton's
surface of 34-35 K, 3 K lower than the canonical value of 38 K derived
from other Voyager 2 experiments (Tyler et al., Science, 1989; Conrath
et al., ibid.; Broadfoot et al., ibid.; Nelson et al., Science, 1990).
Although small, this temperature discrepancy translates into a surface
pressure of N2 an order of magnitude smaller than the measured surface
pressure, 13-19 microbars (Tyler et al., .Science, 1989), assuming vapor
pressure equilibrium. The Imaging Science studies have hypothesized that
a low emissivity for the N2 ice on the surface could bring the various
estimates of surface temperature into agreement.

We have applied Hapke theory to the problem of estimating the
emisssivity of a surface composed of N2 ice. The single scattering
albedo of `particles' composing the surface is computed using published
absorption coefficients for solid N2 over a range of particle sizes. Due
to the weakness of the far-IR absorptions, multiple scattering dominates
over the single particle phase function, but for completeness we explore
the effect of different phase functions. Results for the bolometric
emissivity will be presented and discussed in the context of the above
mentioned studies and a recent re-analysis of the Voyager 2 IRIS data
for Triton (Stansberry et al., this conference).

30.06 Stansberry J. A.* Spencer J. R. Pearl J. C.
Triton's Temperature and Emissivity: Voyager IRIS Data

The original analysis of Voyager 2 infrared observations of Triton
(Conrath et al., Science, 1989) was based on 16 IRIS dayside spectra,
and concluded that Triton's temperature was 38^+3(sub)-4 K. IRIS
actually obtained 75 other spectra, equal in quality to those in the
original study. The entire set of 91 spectra cover parts of Triton's day
and night sides, primarily clustered near the sunward and anti-sunward
longitudes. The dataset includes portions of the south polar cap, the
equatorial bright frost, equatorial dark regions, and the area north of
the northern arctic circle.

The spectrum obtained by averaging all 91 individual spectra was fit
with gray-body irradiance curves for a range of temperatures and
emissivities. The resulting x^2 surface displays a well defined valley
covering a broad range of emissivities (0.1 to unity). The best fit
blackbody temperature is 34.5 K. Such a low temperature is barely within
the stated errors for the previous IRIS analysis, and is significantly
lower than the temperature inferred from the atmospheric pressure. The
Voyager 2 RSS results (Tyler et al., Science, 1989) constrain the
surface pressure to be in the range 13-19 microbars, with a
corresponding range of N2 ice temperature of 37.4' 38 K, assuming vapor
pressure equilibrium. Combining this temperature constraint with our
fits to the average Triton thermal spectrum yields an emissivity range
of 0.15-0.76, using the 90% confidence interval for x2. This result is
consistent with earlier suggestions that Triton's N2 ice has a low
emissivity (Stansberry et al., BAAS, 1990; Stansberry et al., GRL, 1990;
Hillier et al., Science, 1990; Stansberry et al., Icarus, 1992; see also
Pisano et al., this conference), but it should be noted that in this
study we have only constrained the emissivity near 45 micrometers, and
not the emissivity near Triton's blackbody peak at 80 micrometers.

We have also examined subsets of the data, averaging spectra from the
dayside, the nightside, north of the equator, and south of the equator.
Interestingly, the fits to these subsets of the data suggest that the
warmest portions of Triton occurred north of the equator on the
nightside. The best fit blackbody temperaure for that part of the
surface is 38 K, although greybody radiances with low emissivities
(>~0.1) fit the data equally well. We suggest that this is the first
observational evidence for an area on Triton that is devoid of N2 ice.

30.07 Duxbury N. S.* Brown R. H.
Internal Heat Transport for Triton's Geyser-like Plumes

Triton is one of three objects in the Solar System where eruptive
activity has been definitely detected. Since eruptive plumes were
identified on Triton only in the areas of permanent sunlight, the first
explanation for the energy source of the plumes was by the solid-state
greenhouse effect (1). We investigate the plausibility of an alternative
mechanism based on Triton's large internal heat source, estimated in (2)
as 5-20 percent of the amount of Triton's absorbed insolation. Following
(3), we assume that at low stress levels and at Triton's temperatures
diffusion creep is the dominant type of solid-state creep, and thus we
treat nitrogen ice as a Newtonian viscous fluid. We model convective
heat and mass transport in Triton's nitrogen ice for a 2-D steady-state
case with upper stress-free and lower rigid boundaries in the permanent
deposits up to 1 km thick. Applying the numerical analysis from (4) for
Triton's conditions yields a 1 K subsurface temperature increase, (in a
1 km thick nitrogen layer with the temperature difference of 25 K )
compared to the surface value, at the depth of ~ 6 m for the case with
convection versus 40 m depth for the pure conductive case. This
temperature rise is enough to double the nitrogen vapor pressure and
thus to drive the plumes at the supersonic exit speed of 125 m/s to the
observed height of 8 km. It is plausible that the 6 m deep vent is
caused by the multiple phase transitions in solid nitrogen (5). This
work was supported by NASA grant. 1. Brown, R. H., et al., Science 250,
431-435 (1990). 2. Brown, R. H., et al., Science, 251, 1465-1467 (1991).
3. Kirk, R. L., Abstract. LPI abstracts. 631 (1990). 4. Kvernvold, O.,
Geophys. Astrophys. Fluid Dynamics, 12, 273-294 (1979). 5. Duxbury, N.
S. and R. H. Brown, Science 261, 748-751 (1993).

Friday, 1:40 - 2:25 Crystal Ballroom
J. Veverka, Moderator

30A.01-INV Chapman C. R.*
Ida and Its Satellites: What Do They Imply for Asteroid

No abstract available.

SESSION 31 ....... Asteroids III
Friday, 2:55 - 4:35 Crystal Ballroom A
R. P. Binzel and M. Fulchignoni, Moderators

31.01 Binzel R. P.* Bus S. J. Xu S.
Physical Studies of Small Asteroid Families

Advances in the determination of proper elements and family
identifications by researchers such as A. Cellino, Z. Knezevic, A.
Lemaitre, A. Milani, J. Williams, and V. Zappala have led to the
recognition of numerous new asteroid families. Most of these new
families consist of small asteroids, typically having estimated
diameters below 20 km. To investigate the physical properties of these
small asteroid family members, we are using the proven capabilities of
our CCD spectrograph on the 2.4-m and 1.3-m telescopes of the
Michigan-Dartmouth-MlT Observatory located at Kitt Peak, AZ. To date, we
have measured the visible wavelength spectral properties for multiple
members within the Astrid, Dora, Gefion, Hoffmeister, and Pallas
families. We will report on the results of these measurements and their
implications on whether it is likely that these proposed groupings
represent true physical associations.

31.02 Fulchignoni M.* Birlan M. Barucci M. A.
An Extension of the Barucci's Asteroid Taxonomy

Barucci et al. (1987, Icarus 72, 304) used the G-mode statistics to
classify 438 asteroids described by all 7 ECAS colors and IRAS albedos.
An extension of the method allow us to use the taxnomic classes defined
in Barucci et al. to classify the asteroids for which the complete
variables set is not determined. We described the criterion used to
assign an asteroid to a given taxonomic group, which is based on a
statistical test of appurtenance of the sample(the asteroid) to one of
the asteroid classes defined in the Ba taxonomy. To test the method we
used as input each asteroid used to define the taxonomy: the method is
able to assign each asteroid to the right group (i.e. the one to which
it belonged).When the complete set of variables used in the basic
classification is not available for a given asteroid, the method gives
an indication the appurtenance of the taxonomic asteroid to one or more
types, whose the content in the available variables is closer to that
asteroid: lower the number of available variables, higer the number of
possible taxonomic classifications. An attribution of an asteroid to
more than one taxonomic class became possible. The limits of the method
are discussed and the classification of 151 new asteroids is given.

31.03 Dotto E.* Barucci M. A. Fulchignoni M.
Asteroid Lightcurve Inversion Technique: A Model for 1620

To investigate on the model shape of the asteroid 1620 Geographos we
used the lightcurve inversion method developed in 1988 by Fulchignoni
and Barucci (BAAS, 20, 866) and already applied to the asteroid targets
of Galileo mission (Barucci et al., 1992 A&A 266; Dotto et al., 1994 PSS
in press.).

The used numerical algorithm allows us to obtain synthetic lightcurves
by simulating the rotation of the asteroid in correspondence with
several phases, aspects, and obliquity angles and using different
scattering laws. An automatic procedure modifies the parameters
describing the shape and the large scale surface structures of the
asteroid, to minimize the difference between the observed lightcurves
and the synthetic ones. The final result is the model which, using the
minimum number of free parameters, gives synthetic lightcurves best
fitting the observed ones.

In order to obtain a model for Geographos we considered the wall sample
of available lightcurves (Magnusson et al., 1994, in preparation)
observed from 1969 until 1994. The high amplittude of the most part of
the observed lightcurves suggested that Geographos is a very elongated
body, while the sharpness of the principal minimum suggested the
presence of curvature variations.

In order to reproduce all the observed lightcurve features, we
considered several models composed by ellipsoids, the union of
semiellipsoids with or without cutting planes and/or albedo spots. The
resulting model of the shape and the large scale surface structures of
Geographos will be presented and discussed.

31.04 Svedhem H.* Iglseder H. Munzenmayer R.
In situ Measurements of Cosmic Dust by the Hiten Spacecraft

The cosmic dust experiment on the Hiten spacecraft detected over 500
impacts during its three years life in space. The first two years
covered orbits with apogees ranging from 300000 km to 1.5 million km and
perigees from a few thousand to 100000 km. The last year was spent in an
elliptic lunar orbit with apolune up to 50000 km until the spacecraft
(intentionally) hit the lunar surface on 10 April 1993.

The dust experiment detected impacting dust particles by measuring the
charges generated by the hypervelocity impacts of the particles onto the
inner walls of the detector. The aperture of the instrument was 100 cm2
and the field of view about 2pi sr. The lower mass limit for detection
is 10^-13 g at a relative velocity of 10 km/s.

The data has been analyzed to show mass distribution, velocity
distribution and flight direction. The dust flux has been correlated to
the known meteor streams to search for temporal increases due to direct
or indirect (lunar ejecta) impacts by meteor stream related dust.

At least two populations with clearly different properties have been
identified: "apex" particles (low angular momentum) and "beta" particles
(solar radiation pressure dominates gravitational effects). Indications
of other groups can be seen but is not statistically proven. Comparison
with fluxes from models and other experiments have been done.

31.05 Baguhl M.* Grun E. Hamilton D. P. Linkert G. Linkert D.
Riemann R.
Flux, Masses and Speeds of Interstellar Dust Particles Sensed
by Ulysses and Galileo

The Ulysses spaceprobe, looping underneath the Sun's south pole, and
Galileo, enroute to Jupiter, are each equipped with identical dust
detectors that measure the mass, speed and impact direction of incident
dust particles.

After Ulysses began its descent beneath the ecliptic in February 1992,
the majority of dust impacts could be clearly identified as of
interstellar origin by their impact directions and speeds. Traveling
inward to a heliocentric distance of 3.2 AU (attained in March 1994),
Ulysses measured an almost constant flux of such particles. More recent
data inward to 2.7 AU (attained in October 1994) will be presented in
this paper. As Ulysses continues to move inward, changes in the flux
will give clues to the importance of sublimation and gravitational
focusing. In particular, we hope to determine how near the Sun
interstellar dust particles can get before they sublimate away. The
mass and speed distribution can get before they sublimate away. The
mass and speed distribution of interstellar dust particles detected to
date will be given as well.

Impacts of interstellar dust particles are also clearly present in the
Galileo data outside a heliocentric distance of 2.8 AU (attained in
spring 1994). The flux and the speed and mass distributions will be
compared to Ulysses results.

31.06 Gebhardt D.* Hamilton D. P. Burns J. A.
The Motion of Tiny Interplanetary Dust according to Ulysses

Both prior and following Jupiter encounter, the Ulysses spacecraft
detected periodic, high-velocity (20 < v < 56 km/s), collimated streams
of tiny (10^-16 < m < 10^-l4 g) particles emanating from the planet. The
six strongest events typically lasted a day and were separated by 25-28
days; five other bursts from smaller particles have now been seen and,
before the encounter, occur at twice the frequency (Grun et al. 1992,
1993; Baguhl et al. 1993). For grains of this size with expected values
of charge, electromagnetic forces exceed Jupiter's gravity; thus
positively charged dust can be flung from the jovian magnetosphere
(Horanyi et al. 1993; Hamilton and Burns, 1993).

Once this dust has left Jupiter's environs, it is accelerated primarily
by the interplanetary electromagnetic field. We study whether variations
in this field determine which grains reach the spacecraft's detector,
hence accounting for the observed periodicity in Ulysses' receipt of
grains. The solar wind's speed v and magnetic field B suffer major
modulations over timescales like the Sun's rotation period (26-29 days);
thus the convected electric field E = -vxB, prinicipally lying normal to
the ecliptic, contains significant power at similar periods. For an,
idealized history, where the acceleration has constant magnitude but
changes sign every 14 days, particles traverse oscillatory paths whose
centerlines drift off the ecliptic; the characteristics of these
idealized trajectories explain many properties of the spacecraft data.

We now use actual Ulysses measurements of solar wind v and B to estimate
the acceleration histories felt by grains during the six-month interval
when dust bursts struck the spacecraft. To test our hypothesis, we
launch particles having an assumed charge but a range of masses and
relative velocities that satisfy detection criteria. These particles are
sent back from the spacecraft through the transformed acceleration
field. If the paths of most grains pierce Jupiter's magnetosphere, we
will consider the model confirmed. Stay tuned!

31.07 Kortenkamp S. J.* Dermott S. F. Jayaraman S. Xu Y.-L.
Grogan K.
Does the Earth's Circumsolar Ring Funnel Asteroidal Material
Onto the Planet?

The structure of the interplanetary dust cloud is being probed to finer
resolution than ever before with the aid of observations obtained by the
Infrared Astronomical Satellite (IRAS) in 1983. Geometrical arguments
have been used to explain some of the observed detail like the zodiacal
dust bands now known to be associated with the Themis and Koronis
asteroid families. Recent numerical simulations have revealed the
presence of a circumsolar ring of asteroidal particles in resonant lock
with the Earth. Strong evidence for this ring appears to have been found
in the IRAS observations of the zodiacal background. During certain
months of each year the epicyclic motion of the Earth brings us closer
to the ring. This being the case, it has been suggested that the ring
acts as a funnel capable of efficiently depositing asteroidal dust
particles in the Earth's upper atmosphere. We investigate this
possibility by studying the impact probability of dust particles with
the Earth as the particles spiral from the asteroid belt towards the Sun
under the influence of Poynting-Robertson light drag. Probabilities are
calculated for particles that become trapped in the above mentioned
resonances and for those that do not.
This work is supported by NASA grants NGT-40015, NAGW 1923
JSC/NAG9-630, and ADP/NRA 93-OSS-05.

31.08 McDonald G. D.*
A Search for Extraterrestrial Amino Acids in a Martian

The Antarctic shergottite EETA 79001 is believed to be an impact-ejected
fragment of the planet Mars. Detection of organic carbon in barbonate
(druse) samples from tnis meteorite by stepped combustion analysis has
been reported (Wright et al. (1989), Nature 340, 220). The source of
this volatile carbon has been suggested to be endogenous Martian organic

We have found that samples of carbonate (druse) and basalt (lithology A)
material from this meteorite contain approximately 1 ppm and 0.4 ppm
amino acids, respectively. The detected amino acids are those commonly
found in proteins and are almost exclusively L-enantiomers, and are thus
terrestrial contaminants. There is no indication of
alpha-aminoisobutyric acid, the most abundant amino acid in the
Murchison metoerite and a marker for extraterrestrial organics. The
amino acid abundance profile of the druse material resembles that of
terrestrial amino acids present in Antarctic ice, suggesting that the
source of the contaminant amino acids may be ice meltwater which infused
the meteorite during its residence in the Antarctic.

The levels of amino acids in EETA 79001 druse are not alone sufficient
to account for the approximately 800 ppm volatile carbon previously
reported in druse samples by Wright et al. However, our estimates of
total terrestrial biogenic carbon in the druse material, based on the
amino acid analyses reported herein and the organic carbon content of
polar ice, show that ice-borne humic substances can account for the
organic carbon reported by Wright et al.

This work was supported by the NASA Specialized Center of Research and
Training in Exobiology at UCSD.

31.09 Pedicino J. R.* Singer R. B. Reid R. J.
Spectral Identification of Orthopyroxene in Mars Meteorite

The Antarctic meteorite ALH 84001 was recently reclassified as a new
type of martian meteorite based on mineralogical and oxygen isotope
evidence [1] Recent findings suggest that the crystallization age of
this sample makes it far older than any martian meteorite to date.
[2,3]. We have measured the reflectance spectrum and the position of the
'1 micron' band in particular, of the sample. We made our measurements
on the PIRL laboratory spectrophotometer.(0.3 to 1.2 microns) The sample
was approximately 0.25 grams. Ihe spectrum of ALH 84001.28 is that of an
orthopyroxene, consistent with work done earlier this year naming ALH
84001 an orthopyroxenite [1]. The most diagnostic feature in the
spectrum is the position of the '1 micron' band. This sample has
reflectance minimum at 0.90 microns, putting it squarely in the
orthopyroxene range, in contrast with other martian meteorites. The
position of the band also gives us reasonably accurate ranges for
percent composition of its end members. The band position is indicative
of a 0% to 8% Ca abundance as well as a 10% to 45% Fe abundance in
proportion to all three end members, with the remainder residing with
the Mg component [4]. Microprobe analysis of this sample puts the Fe
content at between 20% and 24% of the total for Fe, Mg, and Ca [5] This
meteorite is unique in the small martian collection. It will no doubt
provide new insights into martian geologic history. Additional work done
just this past year has determined a cosmic ray exposure age of 16 +/- 1
Ma for the meteorite [2]. This age puts it in space much longer than the
other meteorites presumably originating on Mars. This result requires
another impact event to deliver the samples to the Earth, or some sort
of the disruption of the body during its journey through space. This
second scenario seems less likely due to its cornposition relative to
the SNC's. The most fascinating result to date is the crystallization
age for ALH 84001. Work by Jagout; has constrained the age to be around
4.5 Ga. Assuming that this number holds up there are two clear
possibilities; (a) this meteorite is not from Mars after all, as an
asteroidal meteorite of this age is common, or (b) this meteonte is a
piece of the ancient martian crust that represents a new look into the
geochemical and geologica evolution of Mars. Perhaps the greatest
advantage to spectroscopic analysis of this in situ sample is the remote
identification of examples of like material on or near the surface of
Mars itself. The crystallization age is indicative of a location
somewhere in the ancient martian highlands that cover over half of the
planet. To date, there has bee no spectral identification of
orthopyroxenite on the surface. However, our sampling and resolution is
very low due to the nature of these spectral observations from the
Earth. Future probes sent to orbit Mars and land on the surface will be
more effectivt at identifying any exposures of such rock types. The
Pathfinder lander will be equipped with a multispectral camera. There
are 6 narrow bandpasses from 0.75 to 1.0 micron allowing effective
discnmination and identification of materials like ALH 84001. This
allows for additional study of the ancient crust of Mars using ALH 84001
as representative of a sample potentially studied by a future lander in
situ. Thanks to T Swindle for use of ALH 84001.
References [1] Ant. Met. News.16(3). [2] Swindle,1994 [3] Jagoutz, 1994
[4] Sunshine, 1993 [5] Papike, 1993 [6] Mittlefehldt, 1994.

31.10 Ryan E. V.* Davis D. R.
Asteroid Collisions: The Impact Disruption of Cooled Iron

New data regarding the basic properties of asteroids are necessary to
better understand collisional processes in the solar system. Critically
important, for instance, when making age estimates for asteroids, are
data on impact strengths of materials. While there is a considerable
body of experimental data on the impact strength of silicate/rock and
ice targets, little is known about the impact strengths of actual solar
system materials found in meteorites: iron/nickel, stony-irons, or
carbonaceous chondrites. For example, Belton et al. (1994, submitted to
Science) compare the crater size-frequency distributions on asteroids
Gaspra and Ida in an attempt to deduce ages for these bodies. To do
this, a strength for the body must be assumed, and one of several
interpretations is that Gaspra has a metal-rich composition. Impact
strength data is necessary to determine whether crater distribution
differences on Gaspra and Ida are due to differences in the impactor
population or to inherent strength differences between the two

We carried out a series of impact experiments at high velocities (< 2
km/s) at the Ames Vertical Gun Range (AVGR) to measure the impact
strength of Gibeon iron-nickel meteorites cooled below the
brittle-ductile transition temperature. Previous studies examining the
material properties of Gibeons (Johnson and Remo 1974, JGR 79,
1142-1146) indicated that the transition temperature should be ~200 K in
a dynamic (i.e., impact) fragmentation event. Temperatures of asteroids
in the main belt are around 150 K, below the expected transition. The
meteorite- targets underwent brittle fracture at a temperature of 167 K,
and the resulting fragment size distribution displayed the typical
two-slope power law behavior often observed for rock targets. Plastic
deformation would not have resulted in a size distribution for the
fragments. This result implies that it is possible for large metal-rich
asteroids to have regolith on their surfaces. The impact strength we
derived from the 4 successful shots for the Gibeon material was
approximately 500 times larger than the impact strength derived for
basalt (e.g., 1.5 x 10^9 erg/cm^3). Further, a very preliminary
examination of the ejecta velocities from the disruption of the Gibeon
meteorites suggest that they may be much higher than those measured from
rock target disruptions. This could have a significant effect on
theories related to asteroid family creation.

31.11-P Liou J. C. Zook H. A.
An Asteroidal Dust Ring of Micron-sized Particles Trapped in
the 1:1 Mean Motion Resonance with Jupiter

We have studied the orbital evolution of micron-sized asteroidal dust
particles under the influence of planetary perturbations, radiation
pressure, Poynting-Robertson drag, and solar wind corpuscular drag.
Radiation pressure changes the semimajor axis of a dust particle when it
is released from its large parent body. This change depends on particle
size, orbital elements of the parent body, as well as where this
particle is released. It is found that a significant fraction of
particles with diameters around 2 micrometers are thrown right into the
1:1 mean motion resonance zone with Jupiter when they are released.
These particles are often trapped in this resonance for tens of
thousands of years and eventually get out of the resonance with close
encounters with Jupiter. While trapped, they move with tadpole-type
orbits around the triangular Lagrangian equilibrium points (L4 or L5) or
in horseshoe-type-orbits. Because of ongoing collisional processes in
the asteroid belt, there is a continuous supply of dust particles from
the asteroid belt to the 1:1 resonance region. Therefore there must
exist a dust ring composed of asteroidal particles in a narrow size
range around 2 micrometers locked into the 1:1 f resonance with Jupiter.

This work was performed while one of us (J.C. Liou) held a National
Research Council- NASA/JSC Research Associateship.

31.12-P Jayaraman S. Dermott S. F.
Implications of the Earth's Resonant Dust Ring

Asteroidal dust particles that spiral in towards