Statistical test of the tidal triggering of earthquakes: contribution of the ocean tide loading effect

Abstract

SUMMARY The possibility of tidal triggering of earthquakes is investigated for a data set that contains 988 globally distributed earthquakes with magnitude of 6.0 or larger. We synthesize the theoretical time history of tidal stress change, including both solid earth tide and ocean tide loading, at the earthquake hypocentre, and assign the phase angle of tidal stress at the occurrence time of each earthquake. The stress due to ocean loading is obtained by convolving the global ocean tide distribution by Schwiderski with the Green's functions for surface point-mass load, which were newly computed for the Preliminary Earth Model of Dziewonski & Anderson. By testing the distribution of the phase angle statistically, we found a significant phase selectivity for normal-fault-type earthquakes; the null hypothesis that earthquakes take place randomly irrespective of the phase angle is rejected at the significance level of 0.54 per cent for the cubic stress component (trace of the tidal stress tensor), and 0.58 per cent for the stress component along the tension axis of the earthquake's focal mechanism. The highest population of normal-fault-type earthquakes appears at the time of maximum extensional stress, implying that a decrease in the confining pressure due to the earth tide is responsible for triggering earthquake occurrence. Such a clear phase selectivity is not seen for strike-slip- and thrust-type earthquakes. The fault-type dependence of the earthquake triggering effect suggests that shear stress change on the fault plane is also an essential factor of tidal triggering of earthquakes.