[Previous] | [Session 35] | [Next]
A.P. Showman (Univ. Arizona), I. Mosqueira (NASA Ames)
High-resolution Galileo images reveal a variety of landforms within Ganymede's bright terrain. Little evidence for volcanic vents or flow fronts exists, and in some regions extensional strains exceeding 50% have occurred. This has led to a model where intense crustal deformation destroys pre-existing terrain (``tectonic resurfacing''). Such resurfacing probably requires strains exceeding 30%, however, implying that it occurred over only a small fraction of bright terrain (since global expansion did not exceed a few percent and no evidence for compression has been found). Furthermore, Galileo images show smooth regions (e.g., Sippar Sulcus and parts of Uruk Sulcus) that have undergone little deformation, and recent stereo analyses by Schenk et al. show that these terrains are low-lying and horizontal. Therefore, resurfacing by low-viscosity liquid (i.e., water) still appears necessary to explain much of the bright terrain.
Here we present models for liquid-water volcanism. Because liquid water is denser than ice, the major difficulty is understanding how liquid reaches the surface. Our favored idea invokes the fact that topography --- such as a global set of graben --- causes subsurface (ahydrostatic) pressure gradients that can ``suck'' subsurface water upward into the bottom of topographic lows (graben). As the low areas become full, the pressure gradients disappear and the resurfacing ceases. This explains the observed straight dark-bright terrain boundaries: water cannot overflow the graben, so resurfacing rarely embays craters and other rough topography. Subsurface liquid water must exist for the scenario to work, of course, and is plausibly provided by tidal heating during an ancient orbital resonance. We summarize the regimes under which the scenario succeeds (it fails in ``generic'' scenarios because the melting is too deep, but works if tidal heating is concentrated along shear zones or if episodic spikes of tidal heating occur). The resurfacing must occur before the topography disappears by viscous relaxation, and we will compare timescales for these processes. We also evaluate alternate resurfacing mechanisms, such as pumping of liquid water to the surface by thermal expansion stresses and buoyant rise of water through a silicate-contaminated crust that is denser than liquid water. Both these scenarios have problems, however, leading us to favor the ``topographic pumping'' mechanism.