TY - JOUR

T1 - Frequency-domain waveform inversion using the phase derivative

AU - Choi, Yun Seok

AU - Alkhalifah, Tariq Ali

N1 - KAUST Repository Item: Exported on 2020-10-01

PY - 2013/9/26

Y1 - 2013/9/26

N2 - Phase wrapping in the frequency domain or cycle skipping in the time domain is the major cause of the local minima problem in the waveform inversion when the starting model is far from the true model. Since the phase derivative does not suffer from the wrapping effect, its inversion has the potential of providing a robust and reliable inversion result. We propose a new waveform inversion algorithm using the phase derivative in the frequency domain along with the exponential damping term to attenuate reflections. We estimate the phase derivative, or what we refer to as the instantaneous traveltime, by taking the derivative of the Fourier-transformed wavefield with respect to the angular frequency, dividing it by the wavefield itself and taking the imaginary part. The objective function is constructed using the phase derivative and the gradient of the objective function is computed using the back-propagation algorithm. Numerical examples show that our inversion algorithm with a strong damping generates a tomographic result even for a high ‘single’ frequency, which can be a good initial model for full waveform inversion and migration.

AB - Phase wrapping in the frequency domain or cycle skipping in the time domain is the major cause of the local minima problem in the waveform inversion when the starting model is far from the true model. Since the phase derivative does not suffer from the wrapping effect, its inversion has the potential of providing a robust and reliable inversion result. We propose a new waveform inversion algorithm using the phase derivative in the frequency domain along with the exponential damping term to attenuate reflections. We estimate the phase derivative, or what we refer to as the instantaneous traveltime, by taking the derivative of the Fourier-transformed wavefield with respect to the angular frequency, dividing it by the wavefield itself and taking the imaginary part. The objective function is constructed using the phase derivative and the gradient of the objective function is computed using the back-propagation algorithm. Numerical examples show that our inversion algorithm with a strong damping generates a tomographic result even for a high ‘single’ frequency, which can be a good initial model for full waveform inversion and migration.

UR - http://hdl.handle.net/10754/554370

UR - http://gji.oxfordjournals.org/cgi/doi/10.1093/gji/ggt351

UR - http://www.scopus.com/inward/record.url?scp=84887538762&partnerID=8YFLogxK

U2 - 10.1093/gji/ggt351

DO - 10.1093/gji/ggt351

M3 - Article

SN - 0956-540X

VL - 195

SP - 1904

EP - 1916

JO - Geophysical Journal International

JF - Geophysical Journal International

IS - 3

ER -