Final Activity Report Summary - FAULTMOD (Quantitative understanding of the interseismic evolution of fault properties: numerical modelling using lab-experiments) A better understanding of the rheological properties of fault zones is key to understanding the behaviour of seismogenic zones and the seismic hazard associated with them. Advances in the performance of computers and numerical methods have allowed considerable progress in the modelling of the dynamic rupture propagation in the past 10 years. One limitation of this approach stems from the paucity of the available information regarding the fault properties at the onset of the rupture and the capacity of the interseismic processes to maintain or re-create complexity. This work has therefore focused on finding new ways to constrain the rheology of fault zones, in particular during the interseismic periods. Particular attention was paid to the evaluation and actual use of the measurement errors during laboratory experiments. An innovative probabilistic (Bayesian) modelling framework was designed for this purpose and enabled the propagation of these sources of variability through forward models of seismogenesis. The challenge consisted in 1) designing new techniques for the quantitative analysis of experimental slow rock deformation, and 2) proposing the first physics-based earthquake renewal model (probability density function for the time to failure). The methodology developed in this project proved useful to bridge the gap between experimental data and numerical models of seismogenesis and it could be applied to the integration of other types of observations. Indeed, we also initiated a very promising study on the analysis of dense surface deformation fields obtained from high-resolution optical images as a constraint for source models and models of fault-zone rheology. More importantly, our Bayesian methodology also enabled us to produce simulated times to failure of the same (probabilistic) nature as the basic ingredients of seismic hazard assessment.
