|
The brochure of the John von Neumann Institute for Computing
is available in English and in German. It can be ordered at the NIC secretariat
(nic@fz-juelich.de).
deutsche Broschüre (pdf) | English brochure (pdf)
 Earth and Environment
Most research efforts studying environmental processes center around the
Earth's atmosphere and
oceans. It is clear, however, that our environment is also shaped and
influenced by processes
taking place in more remote areas, like in space and in the interior
of the Earth. The interior of
our planet is a particularly remote area and only its outermost
"skin" is accessible to direct
observations. In spite of its remoteness, the dynamics of the Earth's
interior has a great impact on
our habitat and our civilization.
Earthquakes and volcanic eruptions are prominent surface expressions
of the inner vitality of our
planet. Less prominent but likewise important, the internal dynamics
plays a crucial role for the
creation and distribution of resources, ranging from ore deposits to
oil reserves and ultimately
also for one of our most precious resources, i.e. groundwater.
Imperceptible to human senses, but
possibly vital for human life is the existence of the Earth's
magnetic field, shielding our
environment against cosmic radiation and particles. The origin of the magnetic field lies deep
within the Earth, in the outer core, ranging from a depth of 2900 to 5100 km. The outer core
consists mainly of molten iron. Convection currents within the outer core transform heat into
motion and finally into magnetic energy. The whole machinery forms a planetary dynamo
generating and maintaining the magnetic field of the Earth.
Computers and the methods of computational physics are essential
tools allowing us to unravel
the dynamics of this inaccessible part of the Earth. In order to
capture relevant scenarios, most
models are extremely demanding and thus represent typical
applications for supercomputers.
Technically, these problems usually boil down to solving a set
of nonlinear partial differential
equations, often in parameter domains, which have little in
common with applications in oceans
or the atmosphere. Convection in the Earth's mantle may serve
as an example here. The Earth's
mantle is made of rocks, however, over geological time scales
it behaves like a viscous fluid. The
viscosity of the material is extremely temperature-dependent,
it varies over many orders of
magnitude. Fluid dynamics in material with such extreme behavior
is still an open field and many
open questions still exist. In this brochure examples are presented
spanning the full range, from
the Earth's surface to the Earth's core.
Groundwater flow takes place in the porous region beneath the Earth's surface. Especially the
heterogeneity of the subsurface area poses a challenge for models simulating the transport
properties of groundwater flow. In order to reach a satisfactory predictive
power of such models,
the mathematical formulation of the processes involved must be as realistic as possible.
Validation in the field is of key importance, observations, hand in hand with numerical
simulations, lead to further insights and to powerful predictions.
Studies on the dynamics of the Earth's interior operate on a more global
scale. The Earth's surface
is split into several plates. Convective motion in the Earth's mantle,
i.e. a region at a depth of
between 100 and 2900 km, is the reason behind the motion of the plates.
Plate tectonics seems to
be a phenomenon unique to the Earth. Other planets, like Mars and Venus
are not known to
exhibit plate tectonics.
The work of Walzer aims at a self-consistent simulation of
mantle convection and plate
tectonics. The simulation of a viscous fluid, whose surface behaves like a rigid plate, but
becomes fluid-like again once it sinks back into the mantle, is known to be a formidable
numerical problem. Increasing computer power has recently facilitated significant advances in
this field.
Besides modeling plate tectonics, efforts are being undertaken towards a
more realistic treatment
of the thermodynamics of the mantle. Incorporating more complex formulations
of the equation
of state of the mantle material will lead to a better understanding of heat
transport processes in
the Earth's interior.
Research on the geodynamo problem may at present be viewed as the "Holy Grail" of
computational geophysics. Geophysicists have only been able to produce self-consistent dynamos
in numerical models for less than ten years. All existing models are still
unrealistic with respect
to the expected parameter values in the Earth's core. But efforts are
progressing. In particular,
further technical developments are necessary to permit the solution of the
magnetohydrodynamics equations describing the evolution of momentum, temperature and
magnetic field strength under the extreme conditions in the Earth's core.
At the moment it seems
difficult to envisage that the next generation of computers, or the one
after next, will allow a
simulation under real core conditions, but experience tells us a story
of unparalleled progress in
this field.
(Ulrich Hansen, Institute for Geophysics, University of Münster)
In Jülich, a software package has been developed with which the groundwater flow and the
transport of contaminants in groundwater can be predicted. The software package originally
consisted of two independent components: "TRACE" computes the water flux in the subsurface.
Based on the results of "TRACE", "PARTRACE" computes the transport of contaminants. The
software package was recently complemented by "PARCHEM", which includes the computation
of interactions between several contaminants during transport in the groundwater.
The picture shows a simulated heterogeneous groundwater flow velocity
field close to a pumping
well. Regions with high velocities are indicated by small vectors,
regions with low velocities are
identified in black. On the right hand side, the high density of
velocity vectors distinguishes the
forced flow towards the well. Streamlines denote the fate of
substances injected into the flow
velocity field at the grayish plane.
(Harry Vereecken, Institute of Chemistry and Dynamics of the Geosphere, Agrosphere, Research Center Jülich)
Tectonic processes and features like orogenesis, subduction or transform
faults are evidence of a
the dynamic interior of the Earth. Thermally driven convection currents
in the Earth's mantle are
the driving forces for the surface dynamics. Due to the extreme conditions and the large
variability of parameters like the viscosity, a numerical investigation of this fluid dynamics
system is exceedingly demanding in terms of numerical stability and computational power.
Our working group has developed a numerical convection model that is suitable for studying the
convective processes governing the dynamics of the Earth's mantle. By additionally employing
suitable rheological relationships we have been able to describe and investigate the interior
dynamics and surface tectonics self-consistently as a coupled system. The illustration shows a
snapshot of a model calculation. The temperature inside the model volume
is visualized by colors
(blue-cold, red-hot) using a volume-rendering technique. At the front of
the box an upstream of
hot material penetrates to the surface, which is mobilized and transported
to the left (as indicated
by white arrows). The box shown in the figure corresponds to a section of
the Earth's mantle; the
height of the model volume has been chosen so as to match the height of the mantle (approx.
2900 km).
(Ulrich Hansen, Helmut Harder, Alexander Loddoch,
Claudia Stein, Institute for Geophysics,
University of Münster)
The Earth's mantle is a polycrystalline solid that - due to holes in the
crystal lattice - behaves like
a viscous fluid on a geological timescale. The viscosity depends on the
temperature, pressure and
also on the content of readily volatile substances, especially water.
The numerical model
describes the dynamics of a compressible medium in a spherical shell,
which is essentially heated
by radioactive elements. This is a quite realistic assumption for the
Earth's mantle. The central
part of the work is the derivation of a viscosity distribution.
Parameter estimates, as derived from
seismologically observed variables, are included in the computations.
The figure shows the color-
coded viscosity distribution at the Earth's surface; the arrows
indicate the velocity of the solid-
state creep. The piecewise, plate-like movement as is also observed
at the Earth's surface is
noteworthy.
(Uwe Walzer, Institute of Geosciences, University of Jena)
The geomagnetic field is generated deep inside the Earth. Within the molten, electrically
conducting iron core of the Earth, i.e. at a depth of approximately 3000 to 5000 kilometers, a
magnetohydrodynamic dynamo maintains the field. The investigation of such a dynamo process
is one of the great challenges for present geophysics. Of particular interest
are episodic field
reversals of the dominant dipole component observed on geological time scales. The main
challenge in numerical simulations of the geodynamo is how to reach an Earth-like parameter
regime where viscous dissipation is small compared to the Coriolis and magnetic Lorentz forces.
In our working group, we have developed a parallel finite volume method for the numerical
solution of spherical dynamo problems. A small temperature difference
between mantle and core
drives a vigorous chaotic flow in the molten core of the Earth. An
initially small magnetic field is
amplified by induction currents until a statistical equilibrium is
reached. The figure displays an
isosurface of the absolute value of the magnetic field strength vector
within the core. In each
hemisphere, the magnetic field at the core mantle boundary is
concentrated in four flux bundles
corresponding to flow cyclones aligned parallel to the axis of
rotation. A field continuation of the
geomagnetic field to the core mantle boundary reveals a similar configuration. In additional
simulations, we explore the transition to small-scale flows by reducing viscous dissipation.
(Helmut Harder, Stephan Stellmach, Ulrich Hansen, Institute of Geophysics,
University of Münster)
| 
