DS10 - Défi des autres savoirs

Exploring complex aseismic slip beyond typical sizes and durations – AtypicSSE

We develop new ways of detecting and characterizing slow slip episodes that could escape detection by classical methods (GPS-based methods, in particular). To do so, we follow approaches using both geodetic and seismological observables. Our targets are the short / small and the long-lived slow slip events, that can be buried in the GPS noise.

- We have made an original link between engineering seismology and the study of the earthquake cycle, by addressing issues related to the mechanics of the subduction interface. Using complementary methods to analyze the frequency content of earthquakes on the Japanese subduction offshore Honshu and Hokkaido, our results indicate that earthquakes frequency contents evolved both spatially and temporally. The spatial variability of the frequency content of interface earthquake was compared to gravity anomalies and interseismic coupling maps, and allowed us to refine the segmentation of the Japanese subduction zone. Focusing on the signal preceding 2011 Tohoku megathrust earthquake, we obtained a very exciting signal before the mainshock, providing new insights on the mechanisms of loading of large earthquakes (Pina-Valdes et al., JGR 2018).

- By analysing GPS time series in Honshu, we could evidence acceleration signals, associated with changes in the state of coupling of the subduction interface before the 2011 Tohoku Earthquake. Such long-term changes had been observed previously offshore Honshu, and our analysis confirms previous findings. In addition to those known signals, we find a long-term decrease of interplate coupling in the area of Boso. This signal is compatible with observations of the dynamics of the seismicity (Marill et al., in prep).

This should extend our view on how (ie, where and when) slow slip occurs in non-tremorgenic subduction zones, both at the short (eg, day) and long (months to years) time scales. We believe that the resolution of such weak SSEs, and hence the accumulation of new slow slip observations, is a pre-requisite to a better understanding of how the seismic and aseismic segments of the subduction interact.

Based on these observations, we will investigate how aseismic slip can trigger earthquakes, and how this triggering depends on depth, duration, migration, periodicity / smoothness (i.e., whether the slow slip occurs at once or over several episodes), and total slip of the SSEs.

Piña-Valdés, J., Socquet, A., & Cotton, F. (2018). Insights on the Japanese subduction megathrust properties from depth and lateral variability of observed ground motions. Journal of Geophysical Research: Solid Earth, 123. doi. org/10.1029/2018JB015743 ?

1. Mathilde Radiguet, Slow slip event and large earthquake interaction along the Mexican subduction zone, invited presentation at JpGU, Japan, 2018.
2. Lou Marill, Anne Socquet, Baptiste Rousset, David Marsan, Aline Deprez, Integrated study of the transient slow slip events beneath the Boso peninsula, Japan, Wegener conference, 2018.
3. B. Gardonio, D. Marsan, A. Socquet, M. Bouchon, J. Jara, Q. Sun, N. Cotte, M. Campillo, Revisiting slow slip events occurrence in Boso Peninsula, Japan, combining GPS data and repeating earthquakes analysis, Wegener conference, 2018.
4. Anne Socquet, Characterizing the long-term preparation of recent megathrust earthquakes, invited presentation at Congresso Geologico Chileno, Concepcion, Chile, Nov. 2018.
5. David Marsan, Earthquakes as a result of nucleating slip, or as the termination of a cascade of smaller ruptures?, Workshop on Avalanche dynamics and precursors of catastrophic events, Les Houches School of Physics, February 2019
6. David Marsan, Maximum earthquake size and seismicity rate from an ETAS model with slip budget, Statsei 11, Japan, August 2019.
7. David Marsan, A stochastic model for repeating earthquakes, Workshop on repeating earthquakes, Paris, July 2019.
8. Lou Marill, Anne Socquet, David Marsan, Exploring aseismic slip complexity on the Japan subduction, Workshop on repeating earthquakes, Paris, July 2019.

Submission summary

Episodes of slow slip are known to occur in various parts of the world, more particularly in subduction zones. They generate transient surface displacements, which can be measured by geodetic observations, typically GNSS. Their seismic signatures can take several forms, from seismicity swarms (at the more shallow depths) to low frequency earthquakes (LFEs) and deep tectonic tremors. They contribute to further load updip locked asperities and might result in the triggering of large earthquakes.
On the subduction zones that have been undergoing a large M8.5+ megathrust earthquake since 2004 (i.e., offshore Sumatra, Chile, and Honshu), the transition between slow slip and seismic rupture is still poorly understood. While instances of slow decoupling episodes have been documented, they are usually associated with regular seismic events, including repeating earthquakes, rather than with tectonic tremors or LFEs. Understanding how the seismic and the aseismic domains interact therefore appears critical towards better evaluating and mitigating seismic risk.

We propose to develop new ways of detecting and characterizing slow slip episodes that could escape detection by classical methods (GPS-based methods, in particular). To do so, we will follow approaches using both geodetic and seismological observables. Our targets are the short / small and the long-lived slow slip events, that can be buried in the GPS noise. Because it is monitored by the best GPS and seismic networks in the world, and has experienced many earthquakes since the 90s, we will focus on the Japanese subduction. We target 2 zones of particular interest, which host slow slip not accompanied by tremor: Pacific plate offshore Honshu, as well as the Philippine Sea plate offshore Boso. The precursory activity leading up to the Tohoku earthquake (2011, M9) will be particularly investigated, as it mixes both long and short term episodes of decoupling, and has a complex slow slip history.

The GSI (Geospatial Information Authority of Japan) GPS data set has been recently acquired by ISTerre and fully processed in double difference using the state of the art technique. Obtained time series exhibit a lower level of noise than those publicly available. Common modes will be removed from time series to reduce the scatter. We will also remove regional non-tectonic loadings as modelled from GRACE. Time series will be analysed using a kinematic approach to invert for the main phases of the seismic cycle. In a second time, we will analyse residual time series in order to extract small slip complexities, through a template matching procedure.

Most of the seismologic and geodetic data are already stored in ISTerre. Seismic data of the JMA (Japan Meteorological Agency) earthquake catalogue, as well as event-based and continuous streams from the Hi-Net network network (High Sensitivity Seismic Network Japan) operated by NIED, will be used to study the seismicity dynamics and repeating earthquakes. Accelerometer data of KiK-NET and K-NET networks will be used to study the ground motions and the time space variations of the frequency content of interface earthquakes.

The results issued from these disciplinary methodological developments will be cross-compared and analysed in the light of mechanical models addressing interactions between aseismic slip and seismic rupture at the scale of a few asperities. These observations will also be integrated into a macroscopic mechanical model of the seismic cycle, in order to better understand the link between creep and seismicity rate. This integrated view of the relations between seismic and aseismic processes will provide us with a model of the seismogenic zone, and a better understanding of the processes occurring at the transition between steady creep and stick slip behavior, in non-tremorgenic subduction zones where this transition is poorly understood.

Project coordination

Anne SOCQUET (isterre)

The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.



Help of the ANR 256,796 euros
Beginning and duration of the scientific project: December 2017 - 48 Months

Useful links

Explorez notre base de projets financés



ANR makes available its datasets on funded projects, click here to find more.

Sign up for the latest news:
Subscribe to our newsletter