Forschungsbericht 1999-2000 | |
Institut für Planetologie
Wilhelm-Klemm-Str. 10 48149 Münster Tel. (0251) 83-33496 Fax: (0251) 83-36301 e-mail: cormack@uni-muenster.de WWW: http://ifp.uni-muenster,.de Geschäftsführender Direktor: Prof. Dr. Elmar K. Jessberger | |
Forschungsschwerpunkte 1999 - 2000
Fachbereich 14 - Geowissenschaften Institut für Planetologie Planetenphysik (Prof. Dr. Tilman Spohn) | ||||
Convection and Differentiation in Icy Satellites
In this work we discuss differentiation scenarios for a convecting ice- rock mixture in order to
understand and model the post-accretional evolution of the Galilean satellite Callisto. The
dimensionless moment of inertia (MoI) factor of Callisto measured by the Galileo spacecraft
and the striking difference in the surface structures between Ganymede and Callisto pose an
intriguing problem to planetologists. The value of 0.359 ± 0.005 for
Callisto's MoI-factor is significantly smaller than the value of 0.4, the value of a constant
density body, but it is too large for a 2- or 3-layer model with the major components ice,
silicate and iron. While the observations for Ganymede suggest a completely differentiated
3-layer structure, Callisto seems to be only partially differentiated. We propose that Callisto's
present state is a snapshot of a slow unmixing process. Compositional and thermal buoyancy
forces drive the convection of the ice-rock mixture in Callisto. In the course of this bulk solid
state convection a gradual unmixing of the two solid phases takes place, leading to a slow
differentiation.
We examine thermal-compositional convection of an ice-rock mixture by numerically solving
the Boussinesq Approximation of the One-Field equations. First, we solve the equations in two
Cartesian coordinates and examine the influence of rock particle size and heterogeneities in the
initial rock distribution. For rock particles with a non-negligible Stokes sinking velocity an
unmixing front develops with zero rock concentration above. The front sinks with the speed of
the rock particles independent of the convection velocity. For rock particles of 1 to
10 m in diameter the Stokes velocity is large enough such that the unmixing front sinks
a significant distance over the 4.5 Ga of Callisto's evolution. If the main fraction of the
rock is smaller than 1 m in diameter the Stokes velocity is negligible, the rock can be
treated as suspended dust and no differentiation occurs by particles settling. But due to the high
ratio of compositional to thermal buoyancy forces, heterogeneities in the inital rock
distribution lead to a differentiation into a compositional stable layering with increasing rock
volume fraction with depth. Since convection then ceases, heat is removed only by conduction
and the temperature in the ice-rock mixture rises. When the thermal gradients overcome the
stabilizing compositional forces the layering is disrupted by the re-onset of thermal convection
that remixes the interior.
Then, we model the thermal evolution and differentiation of Callisto in spherical,
axisymmetric geometry. The differentiation by heterogeneities in the initial rock distribution
can reduce the MoI-factor to the measured value. But during the pure conduction state the
interior heats up and leads to large scale ice melting in Callisto's interior. This initializes a
"run-away'' differentiation and results in a MoI-factor that is too small. Thus, we can rule out
this type of differentiation for Callisto.
For rock particles in the diameter range of 5 to 7 m we find that the differentiation
process is so slow that the MoI-range of Callisto is reached after several Ga. If all rock
particles have diameters of 12 m or more the differentiation is about one order of
magnitude too fast. But by splitting up the rock fraction into one half suspended dust and the
other half are rock particles with diameter of 12 m we obtain another model showing a
slow unmixing process that reaches Callisto's MoI-range after several Ga. We conclude that we
can explain Callisto's current partially differentiated state by a slow sinking unmixing front, if
the main fraction of the rock particles have diameters of 5 to 7 m or the rock fraction is
split into suspended dust and rock particles of 10 m and more in diameter.
Drittmittelgeber:
Beteiligte Wissenschaftler:
Veröffentlichungen: |
||||
Hans-Joachim Peter