HR: 11:50h
AN: S42A-07    [Abstracts]
TI: A Multiscale Approach to Wave-Equation Tomography
AU: * de Hoop, M V
EM: mdehoop@math.purdue.edu
AF: Center for Computational and Applied Mathematics Purdue University, 150 N. University Street, West Lafayett, IN 47907 United States
AU: van der Hilst, R D
EM: hilst@mit.edu
AF: Department of Earth, Atmospheric, and Planetary Sciences M.I.T, 77 Massachusetts Avenue, Cambridge, MA 02139 United States
AB: Both in active source and earthquake seismology, an exciting challenge is to extract more - and more accurate - information from large volumes of broad-band three component data recorded at dense receiver networks. Of our particular interest is the detection and characterization of different scales in the medium, which are diagnostic for the physical, chemical, and geological processes that cause them. To go beyond the measurement and analysis of selected phase arrivals we are developing methods for wave equation transmission and reflection tomography. Through its cross correlation criterion, wave equation tomography is intimately connected to the inverse source problem. In the framework of this inverse problem and the associated adjoint state calculation, borrowed from reflection seismology, we discuss the tomographic sensitivity kernels from a multiresolution analysis point of view. The frame of curvelets appears as a natural tool to understand how scales in the model variation are represented in the data. Furthermore, we will establish how this approach fits in with the notion of so-called banana doughnut type kernels. We will show examples with sythetic data.
DE: 3260 Inverse theory
DE: 7200 SEISMOLOGY
DE: 7260 Theory
DE: 7270 Tomography (6982, 8180)
SC: Seismology [S]
MN: Fall Meeting 2005