Universality of GR b-value gradients for different tectonic regimes and inferences on a differential stress dependence

Variations in earthquakes size-distribution (b-value) have a central importance in modern seismology. Starting from the late 60’s, the possible explanation of such variations has been found into the Earth’s crust stress differences, going far from the classical view of a constancy of the b- value itself. In fact, stress is the determinant parameter controlling the faulting mechanisms of earthquakes: compressions (thrust faults) accompany higher stresses with respect to extensive (normal faults) mechanisms, with transcurrences (strike-slip faults) in the middle of them. In this thesis, it is showed that earthquakes size-distribution, stress and faulting styles have a clear and straightforward connection both on global and local scale: if the magnitude of b is inversely related to differential stress (as confirmed by laboratory experiments), which, in turn, depends on faulting styles, b-value is expected to vary systematically. By using classical formulations of Anderson and Mohr-Coulomb, the different behaviors of b-value on tectonic styles as due to stress are modeled in new, unreported ways. A sinusoidal behavior of b-value on the rake angle of the focal mechanisms is thought to be a good first-order model for expressing the dependence on tectonic style. Moreover, using a ternary scheme of the focal mechanisms, a second-order effect of differential stress on b- value for dip-slip mechanisms is detected. Finally, using high-quality local dataset for Southern California, the single dependences of b (inverse linearity on depth and separation on faulting styles) are modeled, firstly on their own and then together in a single, universal model, able to be the best explicator of the physical reality. All the analyses shown in this thesis result in a big improvement for supporting the theory of the variations of b connected to stress differences.