What is the earthquake fracture energy?

The energy budget of an earthquake is one of the main open questions in earthquake physics. During seismic rupture propagation, the elastic strain energy stored in the rock volume that bounds the fault is converted into (1) gravitational work (relative movement of the wall rocks bounding the fault), (2) in- and off-fault damage of the fault zone rocks (due to rupture propagation and frictional sliding), (3) frictional heating and, of course, (4) seismic radiated energy. The difficulty in the budget determination arises from the measurement of some parameters (e.g., the temperature increase in the slipping zone which constraints the frictional heat), from the not well constrained size of the energy sinks (e.g., how large is the rock volume involved in off-fault damage?) and from the continuous exchange of energy from different sinks (for instance, fragmentation and grain size reduction may result from both the passage of the rupture front and frictional heating). Field geology studies, microstructural investigations, experiments and modelling may yield some hints. Here we discuss (1) the discrepancies arising from the comparison of the fracture energy measured in experiments reproducing seismic slip with the one estimated from seismic inversion for natural earthquakes and (2) the off-fault damage induced by the diffusion of frictional heat during simulated seismic slip in the laboratory. Our analysis suggests, for instance, that the so called earthquake fracture energy (1) is mainly frictional heat for small slips and (2), with increasing slip, is controlled by the geometrical complexity and other plastic processes occurring in the damage zone. As a consequence, because faults are rapidly and efficiently lubricated upon fast slip initiation, the dominant dissipation mechanism in large earthquakes may not be friction but be the off-fault damage due to fault segmentation and stress concentrations in a growing region around the fracture tip.