The surface of Mercury is dominated by compressive tectonic structures named lobate scarps (Fig. 1a). These are surface-breaking thrust faults that result from periods of global planetary contraction. Geological analyses of lobate scarps indicate that they recorded a 3 to 5 km shrinkage of the planet. Since contraction is mainly due to cooling, models of Mercury's thermal evolution can be constrained by attempting at reproducing the limited amount of planetary reduction as predicted by lobate scarps. To this end, we used a combination of 1D and 3D models to infer the most plausible evolution scenarios that may have led to the present-day Mercury (Fig. 1b and 1c).
Figure 1: (a) Image taken by the MESSENGER spacecraft of "Discovery rupes", a 400 km long lobate scarp with an average relief of about 2 km (NASA/JHUAPL/CIW). (b) Model of the time-history of radius change based on a 1D simulation carried out with the code Evolution. The dashed-dotted line indicates the evolution of changes in planetary radius. (c) Temperature isosurface at 1930 K for the same simulation of panel b, but conducted in 3D spherical geometry using GAIA.
While most calculations of mantle convection are based on the assumption of constant thermodynamic parameters, quantities such as the coefficients of thermal expansion and conduction are known to vary considerably both with pressure and temperature. As these two parameters affect respectively the way heat is advected by mantle flow and conducted via thermal diffusion, they also affect the overall dynamics of the mantle (Tosi et al., 2013). The decrease of the coefficient of thermal expansion upon compression, for example, can influence the way hot mantle plumes interact with the phase boundary at 660 km depth (Fig. 2a), allowing for the formation of large ponds of hot material that may help explain recent seismological observations of the upper mantle beneath the Hawaiian archipelago (Tosi & Yuen, 2011). The pressure and temperature dependence of thermal expansivity and conductivity also affect the way tectonic plates subduct in the deep mantle. Our simulations indicate that these two parameters play an important role in the way subducted plates buckle and spread in the lower mantle, leading to a significant thickening as predicted by seismological studies (Fig. 2b).
Figure 2: (a) Bent-shaped plume below the 660 km discontinuity from a simulation with pressure-dependent thermal expansivity conducted with GAIA in 2D cylindrical geometry. (b) Viscous buckling and diffusive spreading of a tectonic plate subducted in the lower mantle induced by pressure- and temperature-dependent thermal expansivity and conductivity. The simulation was conducted with YACC in a 2D rectangular geometry.
Our knowledge about Mars has improved significantly over the past 50 years due to the large amount of information provided by Earth-based and space telescopes, orbiter data, and lander missions. The volcanic history of Mars is fairly well understood from the analysis of surface structures, and partial melting is thought to have played a central role in shaping it. Although Mars is small compared to the Earth - its mass is ten times lower and its radius less than a half - it owns the most spectacular volcanic features in the Solar System. High resolution images of martian volcanoes, provided for example by Mars Express, have shown that some of them exhibit extremely recent lava flows as young as 2 Ma. Volcanic outgassing is an extremely important process that strongly influences a planet's climate as well as certain properties of its interior (Grott et al., 2013) and thus exerts substantial control on the conditions for its habitability. Indeed, for Mars, geological structures show evidence of liquid water on the surface in the early evolution of the planet. Using numerical models, either one-dimensional models in which the convective heat transport is parameterized via appropriate scaling laws, or two- and three-dimensional dynamical models in which the full set of conservation equations for a planetary continuum is solved numerically, we investigate the volcanic history of Mars and estimate the amount of water that can be outgassed from the interior of the planet (Fig. 3).
Figure 3: (a) 1D parameterized model from (Morschhauser et al., 2011) showing crustal thickness (solid line), stagnant lid thickness (dashed line) and the partial melt region (shaded area). (b) 3D dynamical model showing the partial melt region in red color.
Another piece of information is offered by the so-called SNC meteorites (Shergottites, Nakhlites, Chassignites). These are igneous rocks, either basalts or basaltic cumulates. They have been found on Earth but are believed to have originated from Mars as they contain traces of a gas with an elemental and isotopic composition very close to that of the Martian atmosphere as measured by the Viking orbiter. Geochemical analysis of the SNC meteorites implies the presence of three to four isotopically distinct reservoirs. At least two of these reservoirs are thought to be located in the martian interior, indicating a heterogeneous mantle. To investigate the formation of geochemical reservoirs in the interior of Mars and their consequence for the thermo-chemical evolution of the planet lateral variations of geochemical species need to be accounted for (Fig. 4). To this end, we use 2D/3D numerical models which can then also be used for a direct comparison of modeling results with data from independent observations, e.g., from spacecraft measurements or meteorite analyses.
Figure 4: (a) Simulation showing three distinct reservoirs one in the crust (red-orange color) and two in the mantle: an upper mantle reservoir altered with time (blue color) and a primitive lower mantle reservoir (green color). (b) simulation showing four reservoirs: in addition to the previous simulation an initial upper mantle reservoir (magenta color) has been formed.
In its early history Mars has also been subjected to heavy bombardment by large meteorites. These large impacts, which have left traces in the form of huge craters on the surface of the planet, are thought to have had a substantial influence on the thermal evolution of the planet's deep interior, including its dynamo, and must have had global consequences for the climate of early Mars. We are working on models that integrate the various effects of very large impacts with the internal dynamics of mantle convection and magnetic field generation by the core dynamo of early Mars.
Ganymede is the biggest moon in our solar system. Its detection dates back to 1610, where Galileo Galilei was the first to observe it via a telescope. Together with Io, Europa and Callisto, Ganymede is one of the four Galilean moons, which are the biggest moons that orbit Jupiter. Ganymede is composed of equal amounts of silicate rock (similar to terrestrial planets) and water ice. Ganymede not only stands out from the other moons because of its size but also because its the only moon in the solar system, which shows signs of an internally produced magnetic field.
There exist various theories about how Ganymede's magnetic field is produced. The most likely one says that the origin for the magnetic field lies in Ganymede's iron-rich core, that probably contains some sulfur. The responsible process occuring in the core is called a dynamo. Similar to dynamos we know from Earth (e.g. bicycle dynamo), kinetic energy is converted into magnetic energy in Ganymede's core. The kinetic energy stems from convective motions of an iron-rich liquid (electrically conducting). These convective motions are driven by the cooling of Ganymede's core. We distinguish two types of convection: thermal and compositional convection. Whereas thermally driven convection usually operates in the core the first few million years of the planetary evolution, compositionally driven convection is most likely the present driving mechanism for the dynamo. From Earth's core we know that the solidification of iron in the center of the core leads to the formation of an inner solid iron-rich core. As soons as the heavy iron solidifies it leaves behind lighter liquid material. On top of this light material lies heavier core liquid. Such a layering is gravitationally unstable and followingly the light material rises to the top of the core and thereby initiates compositional convection, which drives the Earth's core dynamo.
Ganymede's core has much lower pressures (6 to 10 GPa) than Earth's core (136 to 364 GPa). For such low pressures the melting behaviour of the iron-sulfur-mixture in Ganymede's core changes such that iron solidifies first at the top of the core (for comparison: in Earth's core iron first solidifies in the center). The solidified iron is heavier than the surrounding liquid and starts to sink. During the sinking process the iron eventually remelts in deeper core regions where the core temperatures are higher than the melting temperatures. The entire freezing-remelting-process leads to the formation of a gravitationally stable chemical gradient, where the upper core regions are enriched in light sulfur and the deeper core regions in heavy iron. This so called iron-snow regime bears many interesting features that may be relevant for the origin of Ganymede's magnetic field.
Within the framework of a one-dimensional parametrized model of Ganymede's core we investigate the formation of the stable chemical gradient, i.e. the size of the gradient and the time period to grow that gradient across the entire core. The gradient has striking effects on the core dynamics since it impedes upward directed motions of core fluid. Another interesting aspect that might bring further insights to explain the origin of Ganymede's magnetic field, is the region below the one, where iron solidifies. In that region the remelting of iron creates an gravitationally unstable fluid layer that is on top of lighter fluid. Analogous to Earth, where we find the reversed situation, compositional convection as the dynamo driving mechanism might just as well be caused by the unstable layering. If the magnetic field is generated by that kind of process it will be restricted to the time period it needs to grow the solidification zone across the entire core, because then the remelting zone vanishes and an inner core starts to grow. That means, as soon as an inner core is present this process could not be responsible for the generation of Ganymede's magnetic field. By measuring Ganymede's interior structure more precisely future mission could at least partially test the proposed scenrio.