HR: 1340h
AN: S43B-1316    [Abstracts]
TI: Accuracy of the Discontinuous Galerkin Method for Computational Seismology
AU: * Hermann, V
EM: hermann@geophysik.uni-muenchen.de
AF: Department of Earth and Environmental Sciences, Theresienstr. 41, Munich, 80333, Germany
AU: Kaeser, M
EM: kaeser@geophysik.uni-muenchen.de
AF: Department of Earth and Environmental Sciences, Theresienstr. 41, Munich, 80333, Germany
AB: We present a detailed error analysis for the ADER-DG method on tetrahedral meshes. We consider several different aspects responsible for the accuracy of the scheme: the order of the method due to the degree of the approximation polynomials within each element, the spacial sampling, i.e. the mesh width, and the number of propagated wavelengths. These factors affect the deviation of the numerically computed signal from the reference signal. Furthermore, it is crucial to choose an error norm that describes the accuracy of a synthetic seismogram quantitatively and separates amplitude and phase misfits in the time as well as the frequency domain. The results of this accuracy study help to select an appropriate approximation order for a given mesh spacing, which typically is determined by the geometrical complexity of the problem. Especially with respect to p-adaptive large-scale simulations, where the mesh width might be strongly heterogenous and the approximation order of the scheme is allowed to vary locally, we provide guidelines for the choice of the applied approximation to ensure a desired accuracy. Finally, we apply the proposed method to a test problem suggested by the SCEC code validation project and compare the results with those of other well-established schemes.
DE: 0500 COMPUTATIONAL GEOPHYSICS (3200, 3252, 7833)
DE: 0550 Model verification and validation
DE: 3225 Numerical approximations and analysis (4260)
DE: 3285 Wave propagation (0689, 2487, 4275, 4455, 6934)
DE: 7290 Computational seismology
SC: Seismology [S]
MN: 2007 Fall Meeting