Semi-supervised Surface Wave Tomography with Wasserstein Cycle-consistent GAN: Method and Application on Southern California Plate Boundary Region
Machine learning algorithm has been applied to shear wave velocity (Vs) inversion in surface wave tomography, where a set of starting 1-D Vs profiles and their corresponding synthetic dispersion curves are used in network training. Previous studies showed that the performance of such trained network...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Published: |
Wiley
2022
|
| Online Access: | https://hdl.handle.net/1721.1/143431 |
| _version_ | 1826200836812832768 |
|---|---|
| author | Cai, Ao Qiu, Hongrui Niu, Fenglin |
| author2 | Massachusetts Institute of Technology. Department of Earth, Atmospheric, and Planetary Sciences |
| author_facet | Massachusetts Institute of Technology. Department of Earth, Atmospheric, and Planetary Sciences Cai, Ao Qiu, Hongrui Niu, Fenglin |
| author_sort | Cai, Ao |
| collection | MIT |
| description | Machine learning algorithm has been applied to shear wave velocity (Vs) inversion in surface wave tomography, where a set of starting 1-D Vs profiles and their corresponding synthetic dispersion curves are used in network training. Previous studies showed that the performance of such trained network is dependent on the diversity of the training data set, which limits its application to previously poorly understood regions. Here, we present an improved semi-supervised algorithm-based network that takes both model-generated and observed surface wave dispersion data in the training process. The algorithm is termed Wasserstein cycle-consistent generative adversarial networks (Wasserstein Cycle-GAN [Wcycle-GAN]). Different from conventional supervised approaches, the GAN architecture enables the inclusion of unlabeled data (the observed surface wave dispersion) in the training process that can complement the model-generated data set. The cycle-consistency and Wasserstein metric significantly improve the training stability of the proposed algorithm. We benchmark the Wcycle-GAN method using 4,076 pairs of fundamental mode Rayleigh wave phase and group velocity dispersion curves derived in periods from 3 to 16 s in Southern California. The final 3-D Vs model given by the best trained network shows large-scale features consistent with the surface geology. The resulting Vs model has reasonable data misfits and provides sharper images of structures near faults in the top 15 km compared with those from conventional machine learning methods. |
| first_indexed | 2024-09-23T11:42:32Z |
| format | Article |
| id | mit-1721.1/143431 |
| institution | Massachusetts Institute of Technology |
| last_indexed | 2024-09-23T11:42:32Z |
| publishDate | 2022 |
| publisher | Wiley |
| record_format | dspace |
| spelling | mit-1721.1/1434312023-04-14T15:57:53Z Semi-supervised Surface Wave Tomography with Wasserstein Cycle-consistent GAN: Method and Application on Southern California Plate Boundary Region Cai, Ao Qiu, Hongrui Niu, Fenglin Massachusetts Institute of Technology. Department of Earth, Atmospheric, and Planetary Sciences Machine learning algorithm has been applied to shear wave velocity (Vs) inversion in surface wave tomography, where a set of starting 1-D Vs profiles and their corresponding synthetic dispersion curves are used in network training. Previous studies showed that the performance of such trained network is dependent on the diversity of the training data set, which limits its application to previously poorly understood regions. Here, we present an improved semi-supervised algorithm-based network that takes both model-generated and observed surface wave dispersion data in the training process. The algorithm is termed Wasserstein cycle-consistent generative adversarial networks (Wasserstein Cycle-GAN [Wcycle-GAN]). Different from conventional supervised approaches, the GAN architecture enables the inclusion of unlabeled data (the observed surface wave dispersion) in the training process that can complement the model-generated data set. The cycle-consistency and Wasserstein metric significantly improve the training stability of the proposed algorithm. We benchmark the Wcycle-GAN method using 4,076 pairs of fundamental mode Rayleigh wave phase and group velocity dispersion curves derived in periods from 3 to 16 s in Southern California. The final 3-D Vs model given by the best trained network shows large-scale features consistent with the surface geology. The resulting Vs model has reasonable data misfits and provides sharper images of structures near faults in the top 15 km compared with those from conventional machine learning methods. 2022-06-15T15:10:50Z 2022-06-15T15:10:50Z 2020-12-09 Article http://purl.org/eprint/type/JournalArticle https://hdl.handle.net/1721.1/143431 Cai, Ao, Qiu, Hongrui and Niu, Fenglin. 2020. "Semi-supervised Surface Wave Tomography with Wasserstein Cycle-consistent GAN: Method and Application on Southern California Plate Boundary Region." 10.1002/essoar.10505230.1 Creative Commons Attribution 4.0 International license https://creativecommons.org/licenses/by/4.0/ application/pdf Wiley Wiley |
| spellingShingle | Cai, Ao Qiu, Hongrui Niu, Fenglin Semi-supervised Surface Wave Tomography with Wasserstein Cycle-consistent GAN: Method and Application on Southern California Plate Boundary Region |
| title | Semi-supervised Surface Wave Tomography with Wasserstein Cycle-consistent GAN: Method and Application on Southern California Plate Boundary Region |
| title_full | Semi-supervised Surface Wave Tomography with Wasserstein Cycle-consistent GAN: Method and Application on Southern California Plate Boundary Region |
| title_fullStr | Semi-supervised Surface Wave Tomography with Wasserstein Cycle-consistent GAN: Method and Application on Southern California Plate Boundary Region |
| title_full_unstemmed | Semi-supervised Surface Wave Tomography with Wasserstein Cycle-consistent GAN: Method and Application on Southern California Plate Boundary Region |
| title_short | Semi-supervised Surface Wave Tomography with Wasserstein Cycle-consistent GAN: Method and Application on Southern California Plate Boundary Region |
| title_sort | semi supervised surface wave tomography with wasserstein cycle consistent gan method and application on southern california plate boundary region |
| url | https://hdl.handle.net/1721.1/143431 |
| work_keys_str_mv | AT caiao semisupervisedsurfacewavetomographywithwassersteincycleconsistentganmethodandapplicationonsoutherncaliforniaplateboundaryregion AT qiuhongrui semisupervisedsurfacewavetomographywithwassersteincycleconsistentganmethodandapplicationonsoutherncaliforniaplateboundaryregion AT niufenglin semisupervisedsurfacewavetomographywithwassersteincycleconsistentganmethodandapplicationonsoutherncaliforniaplateboundaryregion |