HR: 08:30h
AN: S31E-03    [Abstracts]
TI: Arbitrary-resolution global sensitivity kernels
AU: * Nissen-Meyer, T
EM: tarje@princeton.edu
AF: Princeton University, Dept. of Geosciences Guyot Hall, Princeton, NJ 08544, United States
AU: Fournier, A
EM: Alexandre.Fournier@obs.ujf-grenoble.fr
AF: Universite Joseph-Fourier, Laboratoire de Géophysique Interne et Tectonophysique Observatoire de Grenoble BP 53, Grenoble cedex 9, 38041, France
AU: Dahlen, F
EM: fad@princeton.edu
AF: Princeton University, Dept. of Geosciences Guyot Hall, Princeton, NJ 08544, United States
AB: Extracting observables out of any part of a seismogram (e.g. including diffracted phases such as P_diff) necessitates the knowledge of 3-D time-space wavefields for the Green functions that form the backbone of Fréchet sensitivity kernels. While known for a while, this idea is still computationally intractable in 3-D, facing major simulation and storage issues when high-frequency wavefields are considered at the global scale. We recently developed a new "collapsed-dimension" spectral-element method that solves the 3-D system of elastodynamic equations in a 2-D space, based on exploring symmetry considerations of the seismic-wave radiation patterns. We will present the technical background on the computation of waveform kernels, various examples of time- and frequency-dependent sensitivity kernels and subsequently extracted time-window kernels (e.g. banana- doughnuts). Given the computationally light-weighted 2-D nature, we will explore some crucial parameters such as excitation type, source time functions, frequency, azimuth, discontinuity locations, and phase type, i.e. an a priori view into how, when, and where seismograms carry 3-D Earth signature. A once-and-for-all database of 2-D waveforms for various source depths shall then serve as a complete set of global time-space sensitivity for a given spherically symmetric background model, thereby allowing for tomographic inversions with arbitrary frequencies, observables, and phases.
DE: 3225 Numerical approximations and analysis (4260)
DE: 3260 Inverse theory
DE: 3285 Wave propagation (0689, 2487, 4275, 4455, 6934)
DE: 7260 Theory
DE: 7290 Computational seismology
SC: Seismology [S]
MN: 2007 Fall Meeting