Deep Learning Tomography by Mapping Full Seismic Waveforms to Vertical Velocity Profiles
Type
Conference PaperAuthors
Kazei, Vladimir
Ovcharenko, Oleg

Plotnitskii, Pavel
Peter, Daniel

Zhang, X.
Alkhalifah, Tariq Ali

KAUST Department
Physical Science and Engineering (PSE) DivisionEarth Science and Engineering Program
Earth Science and Engineering
Extreme Computing Research Center
KAUST
Date
2020Embargo End Date
2021-12-30Permanent link to this record
http://hdl.handle.net/10754/668224
Metadata
Show full item recordAbstract
Building realistic and reliable models of the subsurface is the primary goal of seismic imaging. By employing an ensemble of convolutional neural networks (CNNs), we build velocity models directly from seismic pre-stack data and quantify model uncertainties by analyzing all ensemble results. Most attempts are made to infer models as a whole. Here, CNNs are trained to map subsets of seismic data directly into 1D vertical velocity logs. This allows us to integrate well data into the inversion and to simplify the mapping by using the regularity of active seismic acquisition geometries. The presented approach uses neighboring common midpoint gathers (CMPs) for the estimation of individual vertical velocity logs. Trained on augmentations of the Marmousi model, our CNNs allow for the inversion of sections of the Marmousi II and the Overthrust models. Once the ensemble is trained on a particular dataset, similar datasets can be inverted much faster than with conventional full-waveform inversion.Citation
Kazei, V., Ovcharenko, O., Plotnitskii, P., Peter, D., Zhang, X., & Alkhalifah, T. (2020). Deep Learning Tomography by Mapping Full Seismic Waveforms to Vertical Velocity Profiles. EAGE 2020 Annual Conference & Exhibition Online. doi:10.3997/2214-4609.202011980Sponsors
We thank Jan Walda of Uni Hamburg, Andrey Bakulin of Saudi Aramco, members of the Seismic Wave Analysis Group (SWAG) at KAUST for constructive discussions. We are grateful to Saudi Aramco for support. The research reported in this publication was supported by funding from KAUST.Publisher
EAGE PublicationsConference/Event name
EAGE2020: Annual Conference OnlineAdditional Links
https://www.earthdoc.org/content/papers/10.3997/2214-4609.202011980Relations
Is Supplemented By:- [Software]
Title: vkazei/deeplogs: Velocity model building by deep learning. Multi-CMP gathers are mapped into velocity logs.. Publication Date: 2019-06-30. github: vkazei/deeplogs Handle: 10754/668289
ae974a485f413a2113503eed53cd6c53
10.3997/2214-4609.202011980