HR: 1340h
AN: NG43A-0563 [Abstracts]
TI: Adaptive Simulated Annealing Velocity Modeling for Rayleigh Wave Dispersion
AU: * Pei, D
EM: donghong@seismo.unr.edu
AF: Nevada Seismological Lab University of Nevada, 1664 N. Virginia St., MS 174, Reno, NV 89557
United States
AU: Pullammanappallil, S
EM: satish@optimsoftware.com
AF: Optim Inc., 1664 N. Virginia St., MS 174, Reno, NV 89557
United States
AU: Louie, J
EM: louie@seismo.unr.edu
AF: Nevada Seismological Lab University of Nevada, 1664 N. Virginia St., MS 174, Reno, NV 89557
United States
AB:
We first implemented a new forward computation of Rayleigh dispersion curves from 1-d velocity profiles. Based on the
reflectivity method in terms of generalized reflection and transmission coefficients, we compute the phase velocities of
fundamental and higher modes and corresponding eigen-functions for shallow surface-wave dispersion curves. The significant
small of relative traction residuals at the free surface and comparisons with established methods show that our forward
calculation is accurate and stable even for high frequencies. As a second step, inversion of the dispersion curve is realized
using the adaptive simulated annealing (ASA) method, which is a directed Monte Carlo optimization that finds the global
minimum of a non-linear error function. Instead of using uniform probability distributions, ASA uses a new-generation
probability distribution. This distribution justifies an exponential temperature annealing schedule in perturbing models and
guarantees the convergence of the algorithm. The exponential decrease of temperature leads to a quick location of the global
minimum. Tests on both real and synthetic Rayleigh dispersion data sets indicate that our ASA optimization is fast and
accurate.
DE: 3260 Inverse theory
DE: 3265 Stochastic processes (3235, 4468, 4475, 7857)
DE: 3285 Wave propagation (0689, 2487, 4275, 4455, 6934)
DE: 4420 Chaos (7805)
DE: 7255 Surface waves and free oscillations
SC: Nonlinear Geophysics [NG]
MN: Fall Meeting 2005