A computational method for full waveform inversion of crosswell seismic data using automatic differentiation

Cao, Danping; Liao, Wenyuan

doi:10.1016/j.cpc.2014.11.002

Cited by 12 publications

(6 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…If complex later arriving waveforms are considered for inversion, then CAE can be a powerful tool for picking up local structures of the data. Another technique deserving attention is the automatic differentiation technique for implementing FWI [32], and it has the potential to replace the implicit function theorem. Automatic differentiation is a natural combination of neural network and FWI.…”

Section: Conclusion and Discussionmentioning

confidence: 99%

Skeletonized Wave-Equation Refraction Inversion With Autoencoded Waveforms

Han

Chen

Hanafy

et al. 2021

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

We present a method that skeletonizes the first arriving seismic refractions by machine learning and inverts them for the subsurface velocity model. In this study, first arrivals can be compressed in a low-rank sense with their skeletal features extracted by a well-trained autoencoder. Empirical experiments suggest that the autoencoder's 1 × 1 or 2 × 1 latent vectors vary continuously with respect to the input seismic data. It is, therefore, reasonable to introduce a misfit functional measuring the discrepancies between the predicted and the observed data in a low-dimensional latent space. The benefit of this approach is that an elaborated autoencoding neural network not only refines intrinsic information hidden in the refractions but also improves the quality of inversion for a reliable background velocity model. Numerical tests on both synthetic and field data demonstrate the effectiveness of this method, especially in recovering the low-to-intermediate wavenumber parts of the subsurface velocity distribution. Comparisons are made with the other three relevant methods, the wave-equation travel-time (WT) inversion, the envelope inversion, and the full waveform inversion (FWI). As expected, the cycle skipping problem is alleviated due to the reduction of dimensions of data space. This method outperforms the envelope inversion in resolution, and it is no worse than WT. Moreover, there is no need for careful manual travel-time picking with this methodology. In general, this inversion framework provides an extendable strategy to compress any input data for reconstructing high-dimensional physical parameters.

show abstract

Section: Conclusion and Discussionmentioning

confidence: 99%

Skeletonized Wave-Equation Refraction Inversion With Autoencoded Waveforms

Han

Chen

Hanafy

et al. 2021

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

show abstract

“…In step 5, the second parameter l is fixed, so the optimization problem is solved with respect to k. Similarly in step 7, k is fixed and a gradient-based optimization problem is solved with respect to l. L-BFGS is efficient since the inversion of Hessian matrix is approximated using a short history of previous updates Dk in step 5 and Dl in step 7, respectively. A detailed description of L-BFGS algorithm is available in [7]. Ideally, a rigorous proof of the convergence of the algorithm is essential.…”

Section: Jacobi-type Iterative Algorithmmentioning

confidence: 99%

An adjoint-based Jacobi-type iterative method for elastic full waveform inversion problem

Liao

2015

Applied Mathematics and Computation

Self Cite

View full text Add to dashboard Cite

“…Such differentiable mechanism enables solving the forward and inverse problems via automatic differentiation and can be easily combined with DNNs. Differentiable programming has been applied in geophysics, for problems in seismic tomography (Sambridge et al, 2007), geostatistical seismic inversion (Liu & Grana, 2019), crosswell seismic inversion (Cao & Liao, 2015), full waveform inversion (Richardson, 2018;Zhu et al, 2022) and subsurface flow (D. Li et al, 2020;Rath et al, 2006).…”

Section: Liu Et Almentioning

confidence: 99%

Joint Inversion of Geophysical Data for Geologic Carbon Sequestration Monitoring: A Differentiable Physics‐Informed Neural Network Model

et al. 2023

View full text Add to dashboard Cite

Geologic CO 2 sequestration (GCS) is one of the main mitigation strategies to prevent carbon dioxide (CO 2 ) from entering the atmosphere by capturing and injecting it into subsurface geological formations, such as depleted oil reservoirs and deep saline aquifers (Aminu et al., 2017;Metz et al., 2005). To ensure safe and long-term CO 2 storage, geophysical monitoring surveys are periodically acquired to surveil the behavior of the injected CO 2 and make informed decisions for storage management (Davis et al., 2019). Time-lapse (4-D) seismic acquisition is one of the monitoring techniques widely deployed in GCS projects, where a series of seismic surveys are sequentially acquired to quantify changes in reservoir properties (e.g., elastic properties, fluid pressure and saturation) during and after CO 2 injection (Caspari et al.

show abstract

A computational method for full waveform inversion of crosswell seismic data using automatic differentiation

Cited by 12 publications

References 35 publications

Skeletonized Wave-Equation Refraction Inversion With Autoencoded Waveforms

Skeletonized Wave-Equation Refraction Inversion With Autoencoded Waveforms

An adjoint-based Jacobi-type iterative method for elastic full waveform inversion problem

Joint Inversion of Geophysical Data for Geologic Carbon Sequestration Monitoring: A Differentiable Physics‐Informed Neural Network Model

Contact Info

Product

Resources

About