Ganymede and Callisto 461
The viscous relaxation and modification of craters can
inform us about the nature of the subsurface ice during and
after crater formation. Older craters are distinctly shallower,
showing the action of viscous relaxation through time. How-
ever, there is not a continuum of viscously relaxed craters as
one might expect if this process was ongoing at a constant
rate. Instead, it appears that early craters relaxed quickly,
while more recent craters are being preserved in a stiff ma-
terial. This implies that heat flow was higher in the past, al-
lowing warm ice to flow just below the surface early in solar
system history, while more recently the subsurface ice has
become colder and stiffer. There is overlap in size between
central-dome craters, penepalimpsests, and palimpsests,
implying that impacts of similar energy formed all these
morphologies at different times. Palimpsests are found
only in the most ancient terrains, whereas central-dome
craters appear to be relatively young. Again, it appears that
palimpsests formed early when the subsurface was warm
and flowed easily, penepalimpsests record a time when the
ice was cooling, and dome craters have formed more re-
cently in a thicker layer of cold stiff ice. On Ganymede, the
formation of the bright terrain appears to mark an impor-
tant transition in crater morphology, with no palimpsests or
Valhalla-type multiring basins being formed after the for-
mation of bright terrain. Thus, it appears that heat flows
were higher on Ganymede until the period of bright terrain
formation, and Ganymede’s subsurface became colder and
stiffer after that period.
4.2 Distribution of Craters and Surface Ages
Variations in the areal density of impact craters are observed
on Ganymede and Callisto, giving us information about the
population of impactors and the relative ages of different
surfaces. In general, the highest crater densities are found
on the dark terrain of Ganymede and the plains of Callisto.
Bright terrain on Ganymede has a much lower density of
craters than the dark terrain, supporting the view that it
formed substantially later. The only areas on Callisto with
lower crater densities are the interiors of impact craters and
large multiring basins, where the surface age has been reset
by the impact.
Translating the areal density of impact craters into abso-
lute ages of different surfaces on Ganymede and Callisto is a
tricky proposition. On the Moon, this can be accomplished
by correlating areas of varying crater density on the lunar
surface with physical samples of those surface materials that
have been returned to Earth and that can be precisely dated
in the laboratory using radioisotope techniques. Since we
have no surface samples from the Galilean satellites, we
cannot directly date them. In addition, we cannot be sure
that the same population of debris that impacted the Moon
also impacted the Galilean satellites, so it is dangerous to
simply translate crater densities between these two differ-
ent parts of the solar system. In general, it is agreed that
the surface of Callisto and the dark terrain on Ganymede
represent primordial surfaces, formed shortly after the for-
mation of the planets, 4.5 billion years ago. Bright terrain
on Ganymede could have formed shortly after that, or it
could have formed only a billion years ago. The current
best guess from crater statistics is that bright terrain most
likely formed at some time during the middle half of solar
system history, but obtaining an exact age is likely to remain
elusive for a long time.
Since Ganymede and Callisto are tidally locked and al-
ways have the same side facing Jupiter, it is expected that
they should gather more of the debris coming from outside
the Jupiter system on the sides facing forward in their or-
bital motion (the bug on the windshield effect), and thus
there should be more craters on their leading hemispheres
than on their trailing hemispheres. Callisto does exhibit
such an asymmetry in crater density, but the asymmetry on
Ganymede is much weaker. One hypothesis to explain this
is that Ganymede’s outer ice shell has rotated with respect
to Jupiter in the past and has become locked to Jupiter
more recently, whereas Callisto’s surface has always been
locked with respect to Jupiter. Another piece of evidence to
support this view comes from the study of split comets. In
1994, we witnessed the impact of comet Shoemaker–Levy
9 into Jupiter—this comet had been disrupted into a string
of fragments by a close encounter with Jupiter before the
impact. If such a string of comet fragments hit one of the
satellites on its way out of the Jupiter system, it would form
a line of closely spaced impact craters called a catena, and
these are in fact observed on the surfaces on Ganymede
and Callisto. On Callisto, all the catenae are on the Jupiter-
facing hemisphere, as one would expect from the impact
of a comet on its way out of the system after a close brush
with Jupiter. On Ganymede, one third of the catenae are
found on the other hemisphere, which would be impossible
unless Ganymede’s ice shell had rotated in the past.5. Tectonism and Volcanism
The surface record of tectonic and volcanic activity is the
most obvious difference between Ganymede and Callisto.
Most of Ganymede’s surface has been reworked by some
combination of these processes, whereas Callisto’s surface
may be untouched. Next, we separately consider the roles
of tectonism and volcanism in the extensively resurfaced
bright terrain of Ganymede, the marginally resurfaced dark
terrain of Ganymede, and the relatively pristine surface of
Callisto.5.1 Bright Terrain
Bright terrain covers two thirds of Ganymede’s surface and
is composed of a dense network of intersecting and overlap-
ping areas of parallel ridges and troughs, termed grooved