<p>&#160;</p><p>Earthquakes are spectacular natural disasters, with for example the recent disastrous Sumatra and Tohoku-Oki earthquakes (2004 and 2011, respectively). Presently, predicting earthquakes remains one of the biggest societal challenges in natural science. While seismological observations have much improved in recent years, our understanding of earthquake source physics remains limited due to the scarcity of monitored seismic rupture along similar fault systems, making long- or short-time scale predictions impossible. Friction and fracture are the two keys to understanding earthquakes. Laboratory experiments could be a robust solution to study earthquakes under safe and controlled conditions, which is mandatory to understand and compare the details of earthquake source physics. Conversely to common friction experiments conducted at both slow and seismic slip rates, the stick-slip mechanism is associated to the propagation of a rupture front, i.e. the radiation of seismic waves. Using stick-slip as an earthquake analog coupled to a state-of-the-art high frequency acoustic monitoring system, we demonstrated in the past that accelerations recorded in the kilohertz range on centimeter-sized samples were self-similar to the ones one can expect at the kilometric scale for a large earthquake. Based on this laboratory earthquakes catalogue, we highlighted that acoustic and strain measurements can be used to (i) locate and follow seismicity, (ii) estimate the energy budget of laboratory earthquakes, (iii) discriminate the mode of slip and the rupture speed. Lately, using medium scale experiments, we studied the scale dependence of rupture processes. These new results, notably in term of weakening of faulting and energy balance allowed us to initiate a bridge between laboratory earthquakes, fracture mechanics and natural seismicity. We discuss here how these experimental results can be upscaled to natural earthquakes.</p>
A project of creating atlas aftershocks of strong earthquakes
