Proposal for DEISA Extreme Computing Initiative accepted !

Martin Käser's proposal "Seismic wave propagation solutions for realistic 3D media" for the DEISA (Distributed European Infrastructure for Supercomputing Applications) Extreme Computing Initiative is accepted. In collaboration with Michael Dumbser and Josep de la Puente a new high-resolution approach for wave propagation simulation on massively parallel hardware is further developed and applied to large-scale problems in numerical seismology.

DEISA is an EU FP6 Research Infrastructure Project to advance computational sciences in the area of supercomputing in Europe.
The DEISA Extreme Computing Initiative is aiming at leading, ground breaking applications in selected areas of science and technology dealing with complex, demanding, innovative simulations with a label of excellence from at least one national evaluation committee. The initial focus on “Grand Challenge” applications with only little or moderate application enabling work for the DEISA environment has now been expanded to include medium to long term support for important complex application enabling.

Project Title: Seismic Wave Propagation Solutions for Realistic 3D Media

Project Acronym: SEISSOL

Abstract:
Research on the interior structure of the earth and its geophysical properties are mainly based on results of seismology. Today, computer simulations of the propagation of seismic waves represent an invaluable tool for the understanding of the wave phenomena, their generation and their consequences. However, the simulation of a complete, highly accurate wave field in realistic media with complex geometry and geological rheologies is still a great challenge. Therefore, the aim of the proposed project is the improvement and intensive application of the highly accurate and powerful simulation code SEISSOL in order to provide simulations of realistic earthquake scenarios. The code is able to incorporate complex geological models and accounts for a variety of geophysical processes affecting seismic wave propagation, such as strong material heterogeneities, viscoelastic attenuation and anisotropy. Kinematic models of real earthquake rupture processes, geometrically difficult internal and external material boundaries as well as free surface topography can be included. The code is based on the so-called ADER-Discontinuous Galerkin method and has the unique property of achieving arbitrarily high approximation order for the solution of the governing partial differential equations in space and time using three-dimensional tetrahedral meshes. The application of such highly accurate algorithms on massively parallel high performance computer technologies will contribute to the solution of actual problems in numerical seismology and will improve the prediction of peak ground motions caused by strong earthquake events. This way, more precise estimations of local seismic hazard becomes possible. By synthesizing highly accurate accelerograms the decisions of earthquake engineers can greatly be supported when designing earthquake resistant structures and finally will help to optimize the trade-off between safety and cost.