Global seismic energy scaling relationships based on the type of faulting

Abstract. We derived scaling relationships for different seismic energy metrics for earthquakes around the globe with MW > 6.0 from 1990 to 2022. The seismic energy estimations were derived with two methodologies, the first based on the velocity flux integration and the second based on finite-fault models. In the first case, we analyzed 3331 reported seismic energies derived by integrating far-field waveforms. In the latter methodology, we used the total moment rate functions and the approximation of the overdamped dynamics to quantify seismic energy from 231 finite-fault models (Emrt and EO, EU, respectively). Both methodologies provide compatible energy estimates. The radiated seismic energies estimated from the slip models and integration of velocity records are also compared for different types of focal mechanisms (R, reverse; R-SS, reverse–strike-slip; SS, strike-slip; SS-R, strike-slip–reverse; SS-N, strike-slip–normal; N, normal; and N-SS, normal–strike-slip), and then used to derive converting scaling relations among the different energy types. Additionally, the behavior of radiated seismic energy (ER), energy-to-moment ratio (ER/M0), and apparent stress (τα) for different rupture types at a global scale is examined by considering depth variations in mechanical properties, such as seismic velocities, rock densities, and rigidities. For this purpose, we used a 1-D global velocity model. The ER/M0 ratio is, based on statistical t tests, largest for strike-slip earthquakes, followed by normal-faulting events, with the lowest values for reverse earthquakes for hypocentral depths < 90 km. Not enough data are available for statistical tests at deeper intervals except for the 90 to 120 km range, where we can satisfactorily conclude that ER/M0 for R-SS and SS-R types is larger than for N-type faulting, which also conforms to the previous assumption. In agreement with previous studies, our results exhibit a robust variation in τα with the focal mechanism. Regarding the behavior of τα with depth, our results agree with the existence of a bimodal distribution with two depth intervals where the apparent stress is maximum for normal-faulting earthquakes. At depths in the range of 180–240 km, τα for reverse earthquakes is higher than for normal-faulting events. We find the trend EU > Emrt > EO for all mechanism types based on statistical t tests. Finite-fault energy estimations also support focal mechanism dependence of apparent stress but only for shallow earthquakes (Z < 30 km). The slip distribution population used was too small to conclude that finite-fault energy estimations support the dependence of average apparent stress on rupture type at different depth intervals.