Interdependent Dynamics of EArthquake-prone fault Systems – IDEAS

TASK #1 - Variation of frictional properties with temperature
The existing models do not account for the variations that past ruptures may have caused (friction is defined a priori). Therefore its impact on the overall seismic cycle and on seismic hazard assessment remains to be explored. By implementing a new temperature-dependent rate-and-state friction law inside the BICycle code, we aim to quantify its effect on the size, the magnitude, the frequency and the rupture propagation of earthquakes and how it influences postseismic relaxation.

TASK#2 - Evolution. of bulk properties between majors earthquakes
These changes in bulk properties not only impact the seismic failure but also afterslip. However, they remain to be included in numerical models of seismic cycle. Based on these observations, we propose to tackle this problem by combining two modeling strategies. Using the micromechanical code I have developed we will implement a crack healing law that depends on differential stress and temperature. The results will then be translated in terms of a temporal evolution law of elastic moduli as a function of depth, which will be integrated inside BICycle. This will allow to explore the first-order effect of damage on afterslip, the resulting temporal variations of coupling and also the impact on the aftershocks, since the nucleation size not only depends of friction but also on elastic moduli.

TASK#3 - Temporal variation of permeability and fluid pressure with damage
Using the micromechanical code, on top of the crack healing law (task 2), we will relate the permeability to the density of cracks by implementing the mechanical model of Gueguen & Dienes (1989). It will be used to determine the temporal evolution of pore pressure in the fault zone. This will allow to calculate the evolution of fluid pressure in the fault core after an earthquake and the role it may play on seismic and aseismic behavior.

TASK#4 - Case studies/observation and constraints on the models
The models will be tested against their ability to reproduce the geophysical data and to produce structures that can be observed in the field. This will be based on my previous work and ongoing projects (Taiwan, Philippines) and the numerous studies of my colleagues on the Aegean region.
Long-term perspectives

This project aims to illuminate the unknown physical properties and underlying mechanisms that controls fault behavior, for the wide range of slip rate recorded by geophysical networks. The proposed pathway is to take a concerted and systematic approach in understanding the on-fault and off-fault processes at play during seismic events and between major earthquakes. To achieve such goal, it is necessary to have a dual observation/modeling approach to link the study of natural cases to the models based on the microphysical laws. Today, this link is missing in the study of seismic faulting. By relying on the thematic mobility I have developed and the world-recognized expertise of ISTeP’s researchers in field geology, we are now able to tackle this problem. This unique study will bring together knowledge from fracture mechanics, structural geology, laboratory experiments and geodetic analysis of active faults inside one project. It will add to our understanding of earthquake physics and aseismic deformation, therefore providing a much-needed mechanistic interpretation of fault slip behavior. If successful, the results could be of great societal impact by improving our ability to assess seismic hazard and by defining novel research directions, like setting the relevant parameters to document in the field or using geodesy. This project corresponds to the domain 7.2 of the “Plan d’action ANR 2019”. It will fill the gap highlighted by the INSU “prospective nationale Terre Solide 2016-2020” on “powerful tools to model seismic failure along complex structure in heterogeneous media”.

Thereafter, a natural extension to this project will be to implement a theoretical model that incorporates the different developments highlighted above in one signe model that can produce several seismic cycles. This is truly not trivial and implies a thorough computational effort to include low- and high-strain rate dynamics in one single micromechanical model, while keeping a reasonable computational cost. This long term plan will result in the next generation of seismic cycle models, where dynamic simulations are carefully tuned by going back and forth between physical modeling and observations on natural faults. Such calibrated physical models will provide new ways to assess seismic hazard, well beyond relying on statistics of past earthquakes, and will be used to relate field, geodetic and seismological studies. By exploring independently the effect of the different physical factors on the diverse phases of the seismic cycle, this proposal incrementally brings new complexities and therefore is an essential stage for such long-term plan.

L. Jeandet Ribes, M. Y. Thomas, and H. S. Bhat (in revision). Initial stress state for 2D Plane Strain simulation of a strike-slip fault.
J. Jara, L. Bruhat, M. Y. Thomas, S. Antoine, K. Okubo, E. Rougier, A. J. Rosakis, C. G. Sammis, Y. Klinger, R. Jolivet, H. S. Bhat, 2021. Signature of transition to supershear rupture speed in the coseismic off-fault damage zone. Proceedings of the Royal Society A..477:20210364. 20210364. doi: 10.1098/rspa.2021.0364
S. A. M. den Hartog, M. Y. Thomas, and D. R. Faulkner, 2021. How do Laboratory Friction Parameters Compare With Observed Fault Slip and Geodetically Derived Friction Parameters? Insights From the Longitudinal Valley Fault, Taiwan, Journal of Geophysical Research: Solid Earth, v. 126, e2021JB022390. doi: 10.1029/2021JB022390
A. Canitano, M. Godano, and M. Y. Thomas, 2021. Inherited state of stress as a key factor controlling slip and slip mode: inference from the study of a slow slip event in the Longitudinal Valley, Taiwan, Geophysical Research Letters, v. 48. doi: 10.1029/2020GL090278
J. D. B., Dianala, R. Jolivet, M. Y. Thomas, Y. Fukushima, B. Parsons, and R. Walker, 2020. The relationship between seismic and aseismic slip on the Philippine Fault on Leyte Island: Bayesian modeling of fault slip and geothermal subsidence, Journal of Geophysical Research: Solid Earth, v. 125, p2169-9313. doi: 10.1029/2020JB020052

Active fault zones are complex objects with physical properties and slip behavior constantly evolving in response to external mechanical constraints. In the brittle part of the crust, the deformation is rather localized, and the accumulated stresses are released by slip along the fault plane. Geodetic techniques, combined with seismology, have documented the spatial and temporal variability of slip modes at seismogenic depth (0-40 km). Slip rate on faults span a continuum ranging from mm/yr to m/s, and these seismic and aseismic conditions are not necessarily stable over time. Additionally, numerous studies have highlighted the strong coupling existing between the main fault plane and the surrounding medium. They suggest that on top of the “seismic cycle” there is a superimposed “cycle” where the properties of the fault zone evolve according to the sliding dynamics, which in turn influences the mode of deformation. However, the physics of the processes that controls these behaviors, and how it evolves in space and time, remains poorly understood. This severely limits our ability to assess the potential size, magnitude and recurrence of earthquakes on active faults. Therefore, to improve our understanding of active fault zones, seismic/aseismic slip and the evolving physical properties of the bulk must be studied as a unique system of stress accommodation and no longer as two distinct entities.

However, to address this problem, the current numerical models of seismic cycle cannot be used. Deformation in the brittle crust is modeled by two planes, sliding one against the other and whose behavior is controlled by the properties of the interface only. Moreover, such models usually require to attribute constant properties (pressure, temperature, petrology, microstructure), that do not evolve with the deformation. Therefore, by ignoring the contribution of the evolving medium, the complex feedback, as described above, is not taken into account. It calls for a thorough research effort directed towards the development of a new generation of models that includes this entangled dynamic.

With the help from the ANR program, we propose to study the evolution of the thermo-hydro-mechanical (THM) properties as a function of sliding and the counter-impact on the deformation mode. This project will provide a concerted view on fault behavior using a combined observational and theoretical approach. Developing new numerical tools, we will determine the first-order effect of the spatiotemporal variations of temperature, elastic properties and pore pressure by studying them separately. The theoretical development will be tested according to their ability to reproduce field data. For this part of the project, we will highly rely on previously published work and current projects the team-members are involved in (Taiwan, Philippines, Aegean region). Models will also help to identify the relevant parameters to document in the field or using geodesy, thus applying a true back-and-forth approach between numerical models and observations. This unique study will bring together knowledge from fracture mechanics, structural geology, laboratory experiments and geodetic analysis of active faults inside one project. It will add to our understanding of earthquake physics and aseismic deformation, therefore providing a much- needed mechanistic interpretation of fault slip behavior.