Philosophy of the approach:

It is necessary to develop techniques that deal with the complexity of the geophysical systems.
We proceed in three stages:

1. We construct dynamical systems that reproduce elementary Earth-like behaviors from the nature of the interactions and not from the nature of the heterogeneities.
2. We extract from these models control parameters and characteristic scenarios.
3. We extrapolate the results to real geophysical systems in order to provide a statistical description of these systems.

Why this approach in a study of the faulting in brittle rocks?

1. The fracturing process extends on a large range of spatial scales and may produce highly discontinuous phenomena on a large range of temporal scales (e.g. earthquakes, aftershock sequences, creep).
2. No description of the brittle crust is precise enough to approximate the geometry of the fractures and to measure the relative influence of different physical parameters (e.g. stress corrosion, pore-pressure, heat flow, healing).
3. Fracture populations and earthquakes may be described by similar relationships under extremely different geophysical contexts (intra-plate, inter-plate).