452 Encyclopedia of the Solar System
concentrated in the center, whereas Callisto has a more
homogenous structure. Second, Ganymede was found to
have a relatively strong internal magnetic field, creating its
own “mini-magnetosphere” embedded within Jupiter’s vast
magnetosphere. Finally, the interactions of Ganymede and
Callisto with Jupiter’s rotating, tilted magnetic field show
that both satellites exhibit an induced magnetic field in-
terpreted as evidence for an electrically conducting liquid
water ocean beneath their icy crusts.
2. Interiors
2.1 Interior Structures
The ice/rock bulk composition inferred for Ganymede and
Callisto from their densities led to the natural suggestion
that even modest heating from accretion and the decay of ra-
dioactive elements in the rock fraction would melt ice in the
interior and lead to differentiated interiors—that is, a lay-
ered structure with the denser rock and metal constituents
concentrated closer to the center of the satellite with the
ice in the outer layers. Most analyses following theVoyager
mission operated on the assumption that Ganymede and
Callisto had similar differentiated interior structures, but
the data to test this assumption would not come until the
Galileomission.
Determining the interior structure of a planetary object
is intrinsically difficult, particularly from remote observa-
tions alone. Most of the information about the interior of
our Earth, for instance, comes from over a century of study
of seismic data, where waves created by earthquakes travel
deep through the Earth and provide clues to the density
and composition throughout the interior [seeEarth as a
Planet:Surface andInterior]. So far the only other
world for which we have seismic data is the Moon, acquired
with seismometers left by theApolloastronauts [see The
Moon].
An extremely important quantity that can be used to
assess the distribution of mass inside an object is its mo-
ment of inertia, a dimensionless number; a sphere with
uniform density throughout has a moment of inertia of
0.4, with lower values indicating increasing degrees of mass
concentration near the center. The moments of inertia for
Ganymede and Callisto were measured indirectly by the
radio experiment onGalileo, which measured the pertur-
bations of the spacecraft’s trajectory as it flew by the satel-
lites at low altitude. Although perfect spheres with dif-
ferent moments of inertia have identical external gravity
fields, the key to this experiment is that the distribution
of mass in the interior of a satellite does affect the way
its shape is perturbed from a perfect sphere by rotation
and tides. The rotation rates of Ganymede and Callisto, al-
though slow by terrestrial standards (a little over a week
for Ganymede, and over two weeks for Callisto), are still
sufficient to cause a slight equatorial bulge and polar flat-
tening, whereas Jupiter’s strong gravity raises tidal bulges
on the sub- and anti-Jupiter hemispheres. The combination
of these two effects leads to distinctly nonspherical compo-
nents to the external gravity field (in mathematical terms,
the description of the satellites’ gravity in a spherical har-
monic expansion contains significant J 2 and C 22 terms). The
magnitude of these nonspherical terms is dependent on the
degree of internal massdifferentiation,and they are re-
lated directly to the moment of inertia as long as the object
responds to spin and tidal distortion as a fluid would (i.e.,
hydrostatically).
The surprising results of the Galileo tracking experiment
showed that Ganymede and Callisto have distinctly differ-
ent interiors. The derived moments of inertia for both satel-
lites were lower than they would be for bodies of uniform
density, as expected. Ganymede’s measured value of 0.31
is so low that it implies essentially complete separation of
its water ice from the heavier rock and metal. However,
Callisto has a significantly larger moment of inertia, 0.35.
This is small enough to imply some differentiation, but too
large to be compatible with full separation of light and heavy
components. Callisto probably has some significant portion
of its interior composed of a rock–ice mixture.
The measured moments of inertia can be combined with
the values for the mean density, the size, and the proper-
ties of ice and rock under pressure to construct models of
the satellites’ interiors that match all the known quantities.
Figure 3 shows the best current estimates of their internal
structures. Ganymede is shown with a three-layer structure:
a metallic core, a rock mantle, and a deep water ice upper
layer; Callisto is shown with a two layer structure: a large
rock–ice core, with the fraction of dense material increasing
toward the center, and an upper ice-rich layer.
2.2 Internal Oceans
A major question regarding these icy worlds is whether they
possess subsurface oceans of liquid water. This intriguing
possibility was first raised in the early 1970s by planetary
geochemist John Lewis, who pointed out that radioactive
heating of the satellites’ interiors might result in their in-
ternal temperatures reaching the water ice melting point at
some depth below their surfaces. With the satellites’ densi-
ties known, a relatively simple calculation of internal tem-
perature from the heating produced by the decay of radioac-
tive nuclides in the rock fraction (primarily U, Th, and K)
shows that indeed the ice melting point should be reached
about 75 to 100 km below the surface. More detailed cal-
culations are complicated by several additional factors.
The behavior of water as a function of temperature and
pressure is complex. At the surface of the Earth, only nor-
mal low-density ice, known as Ice-I, exists. It floats in liquid
water and melts at 273 K (0◦C, 32◦F). Increased pressure
decreases the melting temperature, but under terrestrial