Quantitative evaluation of relationship between stress change of large earthquakes and activity of low-frequency earthquakes in Northeast Japan

Abstract:

Deep low-frequency earthquakes (LFEs) in Northeast Japan occur actively beneath the active volcanoes at depths of 20–40 km. LFEs radiate low-frequency seismic waves, with most energy at 2–8 Hz, despite their low magnitudes (M < 2). Regarding LFE activity in the Tohoku region, the number of detected LFEs near Zaozan showed a remarkable increase from about two years after the Tohoku earthquake (Kosuga et al., 2017), which is thought to be due to the stress change of the Tohoku earthquake. However, quantitative analysis for the mechanism of activation in LFEs such as Zaozan has not yet been conducted. Therefore, we investigate the relationship between static stress change of large earthquake and LFEs by evaluating the similarity between focal mechanism of LFEs and stress tensors.

First, we determined focal mechanisms of LFEs for 26 volcanic regions in Northeast Japan. We carry out a grid search to estimate the moment tensor components which minimize the residuals between observed and theoretical S/P ratios. To evaluate estimation error, we select reliable solutions based on Bootstrap test. Next, we computed stress tensors using “Coulomb 3.3” program (Toda et al., 2005; Lin et al., 2004) for three fault models: the 2011 Tohoku earthquake, afterslip and the 2008 Iwate-Miyagi earthquake. Finally, we calculate the moving average of the distance between moment tensor of LFEs and stress tensors every three years in each region.

As a result of this analysis, characteristic temporal distributions were obtained in Naruko and Zaozan. In Zaozan, at the same time as the activation of LFE activity, focal mechanisms have changed to be close to stress tensor of the Tohoku earthquake. Focal mechanisms have transitioned to afterslip like since about three years after the Tohoku earthquake. This result indicates that postseismic deformation such as afterslip or viscoelastic relaxation may be important role of activity change. On the other hand, in Naruko, where LFE activity has calmed down after the Tohoku earthquake, two moment tensors were not similar around 2012, but it has gradually changed to stress tensor like mechanism since about 2013. This recovery may suggest that stress change has slowly progressed. In Naruko, mechanism change in response to stress change of the Iwate-Miyagi earthquake was also observed.