Continuous chatter of the Cascadia subduction zone revealed by machine learning

Rouet‐Leduc, Bertrand; Hulbert, Claudia; Johnson, Paul A.

doi:10.1038/s41561-018-0274-6

Cited by 122 publications

(98 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The observed exponential build-up in seismic energy in tremor frequency bands can be explained by an exponential growth of either the size or the slip rate (or both) of the nucleation phase of a slow slip event. In particular, we know from recent work that the same seismic features map accurately to the slip rate in both the laboratory 13,39 and in Cascadia 20 . Therefore, our results suggest that slow slip often begins with an exponential acceleration on the fault, that can be small enough to not be captured in cataloged tremor.…”

Section: Seismic Power Analysis and The Occurrence Of Slow Slip Inmentioning

confidence: 92%

“…Such patterns allow us to estimate key properties of the laboratory fault, including friction on the fault as well as fault displacement rate. In a first effort to generalize these results to a natural fault system, the analysis of slow slip in Cascadia 20 revealed that statistical characteristics of continuous seismic signals can be used to estimate the displacement rate of GPS stations at the surface. These characteristics are related to seismic power, which is analogous to the acoustic power measured in laboratory experiments (we define seismic power as the average of seismic energy per unit of time, i.e.…”

mentioning

confidence: 99%

See 1 more Smart Citation

An exponential build-up in seismic energy suggests a months-long nucleation of slow slip in Cascadia

et al. 2020

Self Cite

View full text Add to dashboard Cite

Slow slip events result from the spontaneous weakening of the subduction megathrust and bear strong resemblance to earthquakes, only slower. This resemblance allows us to study fundamental aspects of nucleation that remain elusive for classic, fast earthquakes. We rely on machine learning algorithms to infer slow slip timing from statistics of seismic waveforms. We find that patterns in seismic power follow the 14-month slow slip cycle in Cascadia, arguing in favor of the predictability of slow slip rupture. Here, we show that seismic power exponentially increases as the slowly slipping portion of the subduction zone approaches failure, a behavior that shares a striking similarity with the increase in acoustic power observed prior to laboratory slow slip events. Our results suggest that the nucleation phase of Cascadia slow slip events may last from several weeks up to several months.

show abstract

Section: Seismic Power Analysis and The Occurrence Of Slow Slip Inmentioning

confidence: 92%

mentioning

confidence: 99%

An exponential build-up in seismic energy suggests a months-long nucleation of slow slip in Cascadia

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…Machine learning methods such as random forest (RF), sparse multinomial regression, support vector machine, and deep neural network (DNN) have started to find wide applications in Earth sciences (Kuwatani et al, 2014;Petrelli & Perugini, 2016;Rouet-Leduc et al, 2019;Ueki et al, 2017;Zhu et al, 2017). In this study, we apply supervised machine learning algorithms, RF, and DNN, to predict the origin of Cenozoic basalts in Northeast China.…”

Section: 1029/2019gl082322mentioning

confidence: 99%

Involvement of Slab‐Derived Fluid in the Generation of Cenozoic Basalts in Northeast China Inferred From Machine Learning

Zhao

Zhang

Geng

et al. 2019

Geophysical Research Letters

View full text Add to dashboard Cite

The origin and involvement of fluid in the generation of Cenozoic basalts in Northeast China are still under debate. Here we apply the machine learning methods of random forest and deep neural network to train models using data sets of global island arc and ocean island basalts. The trained models predict that most Cenozoic basalts in Northeast China are influenced by fluid and that the fluid activity decreases from east to west. The boundary defined by fluid activity coincides with the westernmost edge of the present‐day stagnant Pacific slab determined by seismic tomography and with the geochemical boundary defined by magnesium isotopes. These observations support the view that the fluid involved in the generation of the basalts is controlled by the stagnant Pacific slab instead of driven by the plume induced by the sinking Izanagi Plate.

show abstract

“…Given continuous seismic spectral data as input, the deep learning model outputs an empirical probability that the seismic data contains tectonic tremor. We term this empirical probability “tremorness.” We showed in previous work that the energy of continuously recorded low‐amplitude seismic waves track the smoothed GPS displacement rate well at long timescales (

30 +

days) and short timescales (1 hr) (Rouet‐Leduc et al, ). The energy‐GPS correlation may suggest that tremor is emitted continuously or quasi‐continuously at the plate interface from its slowly slipping portion (Figure a).…”

Section: Resultsmentioning

confidence: 99%

Probing Slow Earthquakes With Deep Learning

Rouet‐Leduc

Hulbert

McBrearty

et al. 2020

Geophysical Research Letters

Self Cite

View full text Add to dashboard Cite

Slow earthquakes may trigger failure on neighboring locked faults that are stressed sufficiently to break, and slow slip patterns may evolve before a nearby great earthquake. However, even in the clearest cases such as Cascadia, slow earthquakes and associated tremor have only been observed in intermittent and discrete bursts. By training a convolutional neural network to detect known tremor on a single seismic station in Cascadia, we isolate and identify tremor and slip preceding and following known larger slow events. The deep neural network can be used for the detection of quasi‐continuous tremor, providing a proxy that quantifies the slow slip rate. Furthermore, the model trained in Cascadia recognizes tremor in other subduction zones and also along the San Andreas Fault at Parkfield, suggesting a universality of waveform characteristics and source processes, as posited from experiments and theory.

show abstract

Continuous chatter of the Cascadia subduction zone revealed by machine learning

Cited by 122 publications

References 35 publications

An exponential build-up in seismic energy suggests a months-long nucleation of slow slip in Cascadia

An exponential build-up in seismic energy suggests a months-long nucleation of slow slip in Cascadia

Involvement of Slab‐Derived Fluid in the Generation of Cenozoic Basalts in Northeast China Inferred From Machine Learning

Probing Slow Earthquakes With Deep Learning

Contact Info

Product

Resources

About