@article{fdi:010088836,
  title = {{M}ulti-station deep learning on geodetic time series detects slow slip events in {C}ascadia},
  author = {{C}ostantino, {G}. and {G}iffard-{R}oisin, {S}ophie and {R}adiguet, {M}. and {D}alla {M}ura, {M}. and {M}arsan, {D}. and {S}ocquet, {A}.},
  editor = {},
  language = {{ENG}},
  abstract = {{S}low slip events ({SSE}s) originate from a slow slippage on faults that lasts from a few days to years. {A} systematic and complete mapping of {SSE}s is key to characterizing the slip spectrum and understanding its link with coeval seismological signals. {Y}et, {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. {H}ere we present a multi-station deep learning {SSE} detector applied blindly to multiple raw (non-post-processed) geodetic time series. {I}ts power lies in an ultra-realistic synthetic training set, and in the combination of convolutional and attention-based neural networks. {A}pplied to real data in {C}ascadia over the period 2007-2022, it detects 78 {SSE}s, that compare well to existing independent benchmarks: 87.5% of previously catalogued {SSE}s are retrieved, each detection falling within a peak of tremor activity. {O}ur method also provides useful proxies on the {SSE} duration and may help illuminate relationships between tremor chatter and the nucleation of the slow rupture. {W}e 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. {A} deep learning detector applied to raw multi-station geodetic time series data from 2007 to 2022 is able to identify slow slip events in the {C}ascadia region and finds that they are each associated with periods of greater tremor activity},
  keywords = {{PACIFIQUE} ; {ETATS} {UNIS} ; {CALIFORNIA}},
  booktitle = {},
  journal = {{C}ommunications {E}arth and {E}nvironment},
  volume = {4},
  numero = {1},
  pages = {435 [13 ]},
  year = {2023},
  DOI = {10.1038/s43247-023-01107-7},
  URL = {https://www.documentation.ird.fr/hor/fdi:010088836},
}