Tensile earthquakes: Theory, modeling, and inversion

Tensile earthquakes are earthquakes which combine shear and tensile motions on a fault during the rupture process. The geometry of faulting is described by four angles: strike, dip, rake, and slope. The strike, dip, and rake define the orientation of the fault normal and the tangential component of the dislocation vector along the fault. The slope defines the deviation of the dislocation vector from the fault. The strike, dip, and rake are determined ambiguously from moment tensors similarly as for shear sources. The slope is determined uniquely and has the same value for both complementary solutions. The moment tensors of tensile earthquakes are characterized by significant non-double-couple (non-DC) components comprising both the compensated linear vector dipole (CLVD) and the isotropic (ISO) components. In isotropic media, the CLVD and ISO percentages should have the same sign and should depend linearly for earthquakes that occurred in the same focal area. The direction of the linear function between the CLVD and ISO defines the velocity ratio nu(P)/nu(S) in the focal area. The parameters of tensile earthquakes can be retrieved from their moment tensors. The procedure yields the angles describing the geometry of faulting as well as the nu(P)/nu(S) ratio in the focal area. The accuracy of the nu(P)/nu(S) ratio can be increased if a set of moment tensors of earthquakes that occurred in the same focal area is analyzed. The calculation of the nu(P)/nu(S) ratio from moment tensors is an auspicious method which might find applications in tomography of the focal area or in monitoring fluid flow during seismic activity. If the nu(P)/nu(S) ratio is found and well constrained, the parameters of tensile earthquakes can be inverted directly from observed data using a constrained nonlinear inversion. In this inversion, the parameter space can be limited by fixing the nu(P)/nu(S) ratio or forcing the nu(P)/nu(S) ratio to lie within some physically reasonable limits.