Fault on-off versus strain rate and earthquakes energy

We propose that the brittle-ductile transition (BDT) controls the seismic cycle. In particular, the movements detected by space geodesy record the steady state deformation in the ductile lower crust, whereas the stick-slip behavior of the brittle upper crust is constrained by its larger friction. GPS data allow analyzing the strain rate along active plate boundaries. In all tectonic settings, we propose that earthquakes primarily occur along active fault segments characterized by relative minima of strain rate, segments which are locked or slowly creeping. We discuss regional examples where large earthquakes happened in areas of relative low strain rate. Regardless the tectonic style, the interseismic stress and strain pattern inverts during the coseismic stage. Where a dilated band formed during the interseismic stage, this will be shortened at the coseismic stage, and vice-versa what was previously shortened, it will be dilated. The interseismic energy accumulation and the coseismic expenditure rather depend on the tectonic setting (extensional, contractional, or strike-slip). The gravitational potential energy dominates along normal faults, whereas the elastic energy prevails for thrust earthquakes and performs work against the gravity force. The energy budget in strike-slip tectonic setting is also primarily due elastic energy. Therefore, precursors may be different as a function of the tectonic setting. In this model, with a given displacement, the magnitude of an earthquake results from the coseismic slip of the deformed volume above the BDT rather than only on the fault length, and it also depends on the fault kinematics.