Multi-station deep learning on geodetic time series detects slow slip events in Cascadia
Abstract Slow slip events (SSEs) originate from a slow slippage on faults that lasts from a few days to years. A systematic and complete mapping of SSEs is key to characterizing the slip spectrum and understanding its link with coeval seismological signals. Yet, SSE catalogues are sparse and usually...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-11-01
|
| Series: | Communications Earth & Environment |
| Online Access: | https://doi.org/10.1038/s43247-023-01107-7 |
| _version_ | 1797451324308062208 |
|---|---|
| author | Giuseppe Costantino Sophie Giffard-Roisin Mathilde Radiguet Mauro Dalla Mura David Marsan Anne Socquet |
| author_facet | Giuseppe Costantino Sophie Giffard-Roisin Mathilde Radiguet Mauro Dalla Mura David Marsan Anne Socquet |
| author_sort | Giuseppe Costantino |
| collection | DOAJ |
| description | Abstract Slow slip events (SSEs) originate from a slow slippage on faults that lasts from a few days to years. A systematic and complete mapping of SSEs is key to characterizing the slip spectrum and understanding its link with coeval seismological signals. Yet, SSE catalogues are sparse and usually remain limited to the largest events, because the deformation transients are often concealed in the noise of the geodetic data. Here we present a multi-station deep learning SSE detector applied blindly to multiple raw (non-post-processed) geodetic time series. Its power lies in an ultra-realistic synthetic training set, and in the combination of convolutional and attention-based neural networks. Applied to real data in Cascadia over the period 2007–2022, it detects 78 SSEs, that compare well to existing independent benchmarks: 87.5% of previously catalogued SSEs are retrieved, each detection falling within a peak of tremor activity. Our method also provides useful proxies on the SSE duration and may help illuminate relationships between tremor chatter and the nucleation of the slow rupture. We find an average day-long time lag between the slow deformation and the tremor chatter both at a global- and local-temporal scale, suggesting that slow slip may drive the rupture of nearby small asperities. |
| first_indexed | 2024-03-09T14:53:02Z |
| format | Article |
| id | doaj.art-30639c3c4ddb47d5bd11db562bfab9ef |
| institution | Directory Open Access Journal |
| issn | 2662-4435 |
| language | English |
| last_indexed | 2024-03-09T14:53:02Z |
| publishDate | 2023-11-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Communications Earth & Environment |
| spelling | doaj.art-30639c3c4ddb47d5bd11db562bfab9ef2023-11-26T14:20:19ZengNature PortfolioCommunications Earth & Environment2662-44352023-11-014111310.1038/s43247-023-01107-7Multi-station deep learning on geodetic time series detects slow slip events in CascadiaGiuseppe Costantino0Sophie Giffard-Roisin1Mathilde Radiguet2Mauro Dalla Mura3David Marsan4Anne Socquet5ISTerre, Univ. Grenoble Alpes, Univ. Savoie Mont Blanc, CNRS, IRD, Univ. Gustave EiffelISTerre, Univ. Grenoble Alpes, Univ. Savoie Mont Blanc, CNRS, IRD, Univ. Gustave EiffelISTerre, Univ. Grenoble Alpes, Univ. Savoie Mont Blanc, CNRS, IRD, Univ. Gustave EiffelGIPSA-lab, Univ. Grenoble Alpes, CNRS, Grenoble INPISTerre, Univ. Grenoble Alpes, Univ. Savoie Mont Blanc, CNRS, IRD, Univ. Gustave EiffelISTerre, Univ. Grenoble Alpes, Univ. Savoie Mont Blanc, CNRS, IRD, Univ. Gustave EiffelAbstract Slow slip events (SSEs) originate from a slow slippage on faults that lasts from a few days to years. A systematic and complete mapping of SSEs is key to characterizing the slip spectrum and understanding its link with coeval seismological signals. Yet, SSE catalogues are sparse and usually remain limited to the largest events, because the deformation transients are often concealed in the noise of the geodetic data. Here we present a multi-station deep learning SSE detector applied blindly to multiple raw (non-post-processed) geodetic time series. Its power lies in an ultra-realistic synthetic training set, and in the combination of convolutional and attention-based neural networks. Applied to real data in Cascadia over the period 2007–2022, it detects 78 SSEs, that compare well to existing independent benchmarks: 87.5% of previously catalogued SSEs are retrieved, each detection falling within a peak of tremor activity. Our method also provides useful proxies on the SSE duration and may help illuminate relationships between tremor chatter and the nucleation of the slow rupture. We find an average day-long time lag between the slow deformation and the tremor chatter both at a global- and local-temporal scale, suggesting that slow slip may drive the rupture of nearby small asperities.https://doi.org/10.1038/s43247-023-01107-7 |
| spellingShingle | Giuseppe Costantino Sophie Giffard-Roisin Mathilde Radiguet Mauro Dalla Mura David Marsan Anne Socquet Multi-station deep learning on geodetic time series detects slow slip events in Cascadia Communications Earth & Environment |
| title | Multi-station deep learning on geodetic time series detects slow slip events in Cascadia |
| title_full | Multi-station deep learning on geodetic time series detects slow slip events in Cascadia |
| title_fullStr | Multi-station deep learning on geodetic time series detects slow slip events in Cascadia |
| title_full_unstemmed | Multi-station deep learning on geodetic time series detects slow slip events in Cascadia |
| title_short | Multi-station deep learning on geodetic time series detects slow slip events in Cascadia |
| title_sort | multi station deep learning on geodetic time series detects slow slip events in cascadia |
| url | https://doi.org/10.1038/s43247-023-01107-7 |
| work_keys_str_mv | AT giuseppecostantino multistationdeeplearningongeodetictimeseriesdetectsslowslipeventsincascadia AT sophiegiffardroisin multistationdeeplearningongeodetictimeseriesdetectsslowslipeventsincascadia AT mathilderadiguet multistationdeeplearningongeodetictimeseriesdetectsslowslipeventsincascadia AT maurodallamura multistationdeeplearningongeodetictimeseriesdetectsslowslipeventsincascadia AT davidmarsan multistationdeeplearningongeodetictimeseriesdetectsslowslipeventsincascadia AT annesocquet multistationdeeplearningongeodetictimeseriesdetectsslowslipeventsincascadia |