About
This project has two main objectives:
- a scientific one, for improving the detection, monitoring, and characterization of the slow, transient mechanical processes involved in the preparation phase of earthquakes ruptures;
- a technological one, to develop and demonstrate the potential of new, high resolution, innovative optical instruments, for contributing to this scientific goal.
These precursory, weak strain signals are usually seen as a special subset of the more general class of “strain transients”, revealing slow mechanical instabilities in the fault system at depth, in a broad range of space and time scales. Only a small fraction of those transients may mark the slow initiation of major seismic ruptures. These weak signals, often depleted from seismic radiation, and still poorly documented and understood, offer a unique observational window on the friction characteristics of fault zones, which is requested for a sound mechanical and probabilistic approach for failure prediction, and hence relates to the major challenges on the physics of earthquake forecasting.
The most commonly reported sources of transient strain are slow slip events (SSE) on fault patches, lasting hours to months, which have been documented since decades, mostly in the deep subduction interplate, but also in other tectonic contexts, and in the shallower crust. They are often synchronous to small amplitude tectonic tremors, but shallow SSEs are mostly coincident to seismic swarms. The moment magnitude of SSEs detected by GNSS goes rarely down to 6.5, and many smaller SSEs are expected to occur in fault zones.
Detecting these smaller and expectedly very numerous SSEs would allow a much more complete characterization of fault friction, populating the large observational gap in the scaling of the standard « slow earthquakes » family. Up to now, very few SSEs have been precursor to large earthquakes, although a slip initiation process might be common before large interplate earthquake, as inferred from the precursory increase of seismicity rate, and from sequences of seismic repeaters. The latter, which will will focus on in the project, are mechanical asperities on a fault surface which break repeatedly, being loaded by aseismic creep on their surrounding surface, at a rate depending on the local creeping rate and on their magnitude, therefore often used for determining the creeping rate of major interplate faults in subduction zones.
In addition to transient slow slip processes, transient pore pressure changes in fault zones are considered as major cause of natural seismic swarms, as inferred from space-time dynamics and diffusion of the seismicity, but GNSS does not resolve their aseismic strain signature. Their direct evidence and quantification arise from induced and triggered seismicity from controlled fluid injection and production.
A first, major barrier for any significant progress in the understanding and mechanical modelling of these strain transients is the scarcity of the relevant, high resolution records close enough to the sources: Dense GNSS and INSAR monitoring only detect the rare largest transients, with moment magnitude Mw> 6-6.5, at the strain level of 10-6 ; and high resolution instruments, like strainmeters and tiltmeters, are much less common in field observatories, if not totally absent. This has several causes: (1), for the boreholes strainmeters, which present the highest resolution of 10-10 (by integrating the strain over up to 3 m long sensing part of the probe), the cost is very high: typically a minimum 200 k€ when including their standard installation in 150-200 m deep boreholes; (2), the tiltmeter boreholes present a fair resolution of 10-8, but also a high sensitivity to local noise in the rock (submetric scale) and a poor long term stability (instrumental drift); (3), long base, hydrostatic tiltmeters have better performances in resolution (10-10) and in long term stability (10-8 radians per month) when installed in tunnels, but are much more difficult to install and are more fragile systems; and (4), common to all these sensors, the difficulty of the extraction of the relevant signals of internal origin, to be separated from the influence of earth and sea tide, from external factors (temperature, air pressure, rain ), and from very local noises in the rock mass; the latter can be however reduced using long base sensors (tiltmeters or strainmeter).
To be safely identified as a strain signal from deep relevant sources, synchronous signals should be recorded on arrays of several, distant stations. This allows to eliminate signals detected on single instruments and produced by local sources or from the instrument itself, unrelated to deep processes. Owing to the fast decay of the strain signal with distance d to the source, as d-3, the distance between station should not exceed much the depth of the sources to allow for redundancy. Most strain monitoring array lack such densities, due to both cost and complexity of installation.
In this scientific and instrumental context of transient strain monitoring in seismically active area, we propose to overcome these barriers taking advantage of the conjunction of:
A privileged access to two sites showing an exceptionally high level of seismic activity, at two very different scales: the western Corinth rift (Greece), which is the region with highest microseismic density and strain rate in the whole Mediterranean, densely monitored by our team (IPGP and ENS) for the last 20 years; and the deep active mine of Garpenberg (Sweden), monitored for the last 6 years by the team (INERIS, coll. IPGP), with the rare opportunity to reach the active fault zone of repeaters.
The outcome of 10 years of R&D of our team (ESEO-IPGP-ENS), focussing on innovative high resolution optical instruments for strain and vibration, which provide specific assets in terms of cost, resolution, and/or stability, with respect to commercial ones: one optical long base hydrostatic tiltmeter, one optical seismometer, and one optical strainmeter, the most recently designed sensor to be developed in the frame of the present project.
The proper design and adaptation of these three sensor type in the monitoring of these two selected sites should considerably improve the detection of strain transients, thus allowing for significant and generic advances on the modelling of the seismic cycle of asperities and on the unstable seismic- aseismic coupling in fault zones. As a complementary “product” of the project, the new instruments will be demonstrated and qualified in the field, and advertised for being used in other contexts of geo-monitoring in academic or industrial environment.
