Unsupervised deep learning with higher-order total-variation regularization for multidimensional seismic data reconstruction
In 3D marine seismic acquisition, the seismic wavefield is not sampled uniformly in the spatial directions. This leads to a seismic wavefield consisting of irregularly and sparsely populated traces with large gaps between consecutive sail-lines especially in the near-offsets. The problem of reconstr...
Published in: | GEOPHYSICS |
---|---|
Main Authors: | , , , , , , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
Society of Exploration Geophysicists Foundation
2021
|
Subjects: | |
Online Access: | http://hdl.handle.net/10852/89335 http://urn.nb.no/URN:NBN:no-91945 https://doi.org/10.1190/geo2021-0099.1 |
id |
ftoslouniv:oai:www.duo.uio.no:10852/89335 |
---|---|
record_format |
openpolar |
spelling |
ftoslouniv:oai:www.duo.uio.no:10852/89335 2023-05-15T15:39:04+02:00 Unsupervised deep learning with higher-order total-variation regularization for multidimensional seismic data reconstruction Greiner, Thomas Larsen Lie, Jan-Erik Kolbjørnsen, Odd Kjelsrud Evensen, Andreas Harris Nilsen, Espen Zhao, Hao Demyanov, Vasily Gelius, Leiv Jacob 2021-11-16T09:41:35Z http://hdl.handle.net/10852/89335 http://urn.nb.no/URN:NBN:no-91945 https://doi.org/10.1190/geo2021-0099.1 EN eng Society of Exploration Geophysicists Foundation NFR/287664 http://urn.nb.no/URN:NBN:no-91945 Greiner, Thomas Larsen Lie, Jan-Erik Kolbjørnsen, Odd Kjelsrud Evensen, Andreas Harris Nilsen, Espen Zhao, Hao Demyanov, Vasily Gelius, Leiv Jacob . Unsupervised deep learning with higher-order total-variation regularization for multidimensional seismic data reconstruction. Geophysics. 2021 http://hdl.handle.net/10852/89335 1954978 info:ofi/fmt:kev:mtx:ctx&ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.jtitle=Geophysics&rft.volume=&rft.spage=&rft.date=2021 Geophysics 1 62 https://doi.org/10.1190/geo2021-0099.1 URN:NBN:no-91945 Fulltext https://www.duo.uio.no/bitstream/handle/10852/89335/2/geo2021-0099.1.pdf 0016-8033 Journal article Tidsskriftartikkel Peer reviewed AcceptedVersion 2021 ftoslouniv https://doi.org/10.1190/geo2021-0099.1 2021-12-01T23:32:32Z In 3D marine seismic acquisition, the seismic wavefield is not sampled uniformly in the spatial directions. This leads to a seismic wavefield consisting of irregularly and sparsely populated traces with large gaps between consecutive sail-lines especially in the near-offsets. The problem of reconstructing the complete seismic wavefield from a subsampled and incomplete wavefield, is formulated as an underdetermined inverse problem. We investigate unsupervised deep learning based on a convolutional neural network (CNN) for multidimensional wavefield reconstruction of irregularly populated traces defined on a regular grid. The proposed network is based on an encoder-decoder architecture with an overcomplete latent representation, including appropriate regularization penalties to stabilize the solution. We proposed a combination of penalties, which consists of the L2-norm penalty on the network parameters, and a first- and second-order total-variation (TV) penalty on the model. We demonstrate the performance of the proposed method on broad-band synthetic data, and field data represented by constant-offset gathers from a source-over-cable data set from the Barents Sea. In the field data example we compare the results to a full production flow from a contractor company, which is based on a 5D Fourier interpolation approach. In this example, our approach displays improved reconstruction of the wavefield with less noise in the sparse near-offsets compared to the industry approach, which leads to improved structural definition of the near offsets in the migrated sections. Article in Journal/Newspaper Barents Sea Universitet i Oslo: Digitale utgivelser ved UiO (DUO) Barents Sea GEOPHYSICS 1 62 |
institution |
Open Polar |
collection |
Universitet i Oslo: Digitale utgivelser ved UiO (DUO) |
op_collection_id |
ftoslouniv |
language |
English |
description |
In 3D marine seismic acquisition, the seismic wavefield is not sampled uniformly in the spatial directions. This leads to a seismic wavefield consisting of irregularly and sparsely populated traces with large gaps between consecutive sail-lines especially in the near-offsets. The problem of reconstructing the complete seismic wavefield from a subsampled and incomplete wavefield, is formulated as an underdetermined inverse problem. We investigate unsupervised deep learning based on a convolutional neural network (CNN) for multidimensional wavefield reconstruction of irregularly populated traces defined on a regular grid. The proposed network is based on an encoder-decoder architecture with an overcomplete latent representation, including appropriate regularization penalties to stabilize the solution. We proposed a combination of penalties, which consists of the L2-norm penalty on the network parameters, and a first- and second-order total-variation (TV) penalty on the model. We demonstrate the performance of the proposed method on broad-band synthetic data, and field data represented by constant-offset gathers from a source-over-cable data set from the Barents Sea. In the field data example we compare the results to a full production flow from a contractor company, which is based on a 5D Fourier interpolation approach. In this example, our approach displays improved reconstruction of the wavefield with less noise in the sparse near-offsets compared to the industry approach, which leads to improved structural definition of the near offsets in the migrated sections. |
format |
Article in Journal/Newspaper |
author |
Greiner, Thomas Larsen Lie, Jan-Erik Kolbjørnsen, Odd Kjelsrud Evensen, Andreas Harris Nilsen, Espen Zhao, Hao Demyanov, Vasily Gelius, Leiv Jacob |
spellingShingle |
Greiner, Thomas Larsen Lie, Jan-Erik Kolbjørnsen, Odd Kjelsrud Evensen, Andreas Harris Nilsen, Espen Zhao, Hao Demyanov, Vasily Gelius, Leiv Jacob Unsupervised deep learning with higher-order total-variation regularization for multidimensional seismic data reconstruction |
author_facet |
Greiner, Thomas Larsen Lie, Jan-Erik Kolbjørnsen, Odd Kjelsrud Evensen, Andreas Harris Nilsen, Espen Zhao, Hao Demyanov, Vasily Gelius, Leiv Jacob |
author_sort |
Greiner, Thomas Larsen |
title |
Unsupervised deep learning with higher-order total-variation regularization for multidimensional seismic data reconstruction |
title_short |
Unsupervised deep learning with higher-order total-variation regularization for multidimensional seismic data reconstruction |
title_full |
Unsupervised deep learning with higher-order total-variation regularization for multidimensional seismic data reconstruction |
title_fullStr |
Unsupervised deep learning with higher-order total-variation regularization for multidimensional seismic data reconstruction |
title_full_unstemmed |
Unsupervised deep learning with higher-order total-variation regularization for multidimensional seismic data reconstruction |
title_sort |
unsupervised deep learning with higher-order total-variation regularization for multidimensional seismic data reconstruction |
publisher |
Society of Exploration Geophysicists Foundation |
publishDate |
2021 |
url |
http://hdl.handle.net/10852/89335 http://urn.nb.no/URN:NBN:no-91945 https://doi.org/10.1190/geo2021-0099.1 |
geographic |
Barents Sea |
geographic_facet |
Barents Sea |
genre |
Barents Sea |
genre_facet |
Barents Sea |
op_source |
0016-8033 |
op_relation |
NFR/287664 http://urn.nb.no/URN:NBN:no-91945 Greiner, Thomas Larsen Lie, Jan-Erik Kolbjørnsen, Odd Kjelsrud Evensen, Andreas Harris Nilsen, Espen Zhao, Hao Demyanov, Vasily Gelius, Leiv Jacob . Unsupervised deep learning with higher-order total-variation regularization for multidimensional seismic data reconstruction. Geophysics. 2021 http://hdl.handle.net/10852/89335 1954978 info:ofi/fmt:kev:mtx:ctx&ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.jtitle=Geophysics&rft.volume=&rft.spage=&rft.date=2021 Geophysics 1 62 https://doi.org/10.1190/geo2021-0099.1 URN:NBN:no-91945 Fulltext https://www.duo.uio.no/bitstream/handle/10852/89335/2/geo2021-0099.1.pdf |
op_doi |
https://doi.org/10.1190/geo2021-0099.1 |
container_title |
GEOPHYSICS |
container_start_page |
1 |
op_container_end_page |
62 |
_version_ |
1766370504765079552 |