Full Waveform Inversion
A joint velocity model building method with Gaussian beam tomography and full waveform inversion for land seismic data
Oil Geophysical Prospecting,2019,Vol 54,No. 05
【Abstract】 The full waveform inversion (FWI) based on the prestack seismic wave field fitting is facing a few challenges in the application, for instances, low signal-to-noise ratio (SNR) land seismic data, missing low frequency information, complicated near-surface, and huge memory needed for massive 3D seismic data processing. In terms of low precision land seismic migrations, in this paper, we propose a velocity model building method by Gaussian beam tomography, which not only makes up the medium wave-number velocity components unachievable in the conventional methods based on ray tomography but also provides FWI with a high precise macroscopic velocity model that helps to effectively avoid cycle skips. Besides, in order to reduce the dependency of inversion on low frequencies, we replace the L 2 norm with a cross-correlation function based on phase matching, which figures out validly the problem of intensive computing of FWI. Furthermore, with a self-adjoint forward operator from pseudo-conservative wave equations, the gradient computing capacity based on the adjoint-state method is enhanced dramatically, which improves the inversion adaptation and realizes high-resolution velocity model building for land seismic data.
Oil Geophysical Prospecting,2018,Vol 53,No. 01
【Abstract】 Real earth media are characterized by the visco-elasticity, so it is necessary to consider the absorption and attenuation effects in the full waveform inversion. In this paper, we propose a visco-acoustic wave full waveform inversion in the 2D time-domain. According to the visco-elastic wave equation based on generalized standard linear solid (GSLS) model, a first-order velocity-stress visco-acoustic wave equation is obtained. Then the corresponding P-wave velocity gradient is derived. Based on high-order staggered-grid finite-difference method and conjugate gradient method, the visco-acoustic full waveform inversion in the 2D time domain based on first-order velocity-stress visco-acoustic wave equation is realized. Numerical examples demonstrate the effectiveness of the proposed method. Compared with other methods without considering the absorption and attenuation effects, the P-wave velocity inversion results with the proposed method are much more accurate.
Oil Geophysical Prospecting,2017,Vol 52,No. 06
【Abstract】 High-precision full waveform inversion depends on both low frequency seismic data and the initial model. However, low-frequency seismic information is usually not available, and an inaccurate initial model will lead to cycle skips. At the same time, inversions might cause multi-solutions and ill-posed problems. As well-logging data are characterized by high resolution and broadband information, we take it as a constraint for full waveform inversion. We correct initial velocity models with reliable low-frequency information extracted from logging data. Then, we extend objective functions by a prior model, where the prior model drives optimization steps as a constraint to minimize the objective functions. Numerical tests on both EAGE overthrust model and real data show that the proposed inversion improves results.
Oil Geophysical Prospecting,2017,Vol 52,No. 06
【Abstract】 We propose in this paper a 3D full waveform inversion method based on reconstructed source wavelet with wave equation. We resolve the wave equation with the direct waves at near offsets of prestack shot gathers as boundary conditions and take the time series at the source point as source wavelet. This method uses the principle of source wavefield reconstruction to simulate the inverse process of the wave propagation along the surface, and extracts direct waves at near offsets, which avoids the interference from the waves such as refraction wave, reflection wave and diving wave. So the method has less dependence on the initial model. For tests, we select the 3D SEG/EAGE overthrust model and field data to analyze the phase difference of 3D seismic wavelet based on the source wavefield reconstruction theory of wave equation. The test results show that there is only phase difference between the reconstructed wavelet and the input wavelet. After the phase adjustment, the reconstructed wavelet is used for the 3D full waveform inversion; the inversion velocity is close to the real one. The results are more reliable at channels. The waveform and phase between the observed records and simulated records have also good correspondence, which validates the feasibility and applicability of the proposed method.
Oil Geophysical Prospecting,2016,Vol 51,No. 02
【Abstract】 Elastic full-waveform inversion (EFWI) is a powerful tool for estimating high-accuracy subsurface elastic parameters. However the nonlinearity in EFWI is much severer than that in acoustic FWI. We propose a multi-step EFWI strategy to mitigate the nonlinearity for the reconstruction of high-accuracy P- and S-wave velocities. We use envelope-based EFWI as first step to update the long wavelength components of P- and S-wave velocities. Using the envelope-based EFWI's results as initial models, a time-domain modeling based frequency-domain EFWI is used to further reduce nonlinearity and recover short wavelength components of the models with hierarchical multi-scale inversion strategy. The numerical test demonstrates that, when the low frequency and long offset components are not available in seismic data, our multi-step strategy is effective for reconstructing high-accuracy P- and S-wave velocities, especially for S-wave velocity.