Spatial and temporal variations of radiated seismic energy estimated for repeating earthquakes in northeastern Japan; implication for healing process
Abstract:Repeating earthquakes are shear slip on the plate interface, and helpful to monitor long-term deformation in subduction zones. Previous studies have measured the size of repeating earthquakes mainly using seismic moment, to calculate slip amount in each event. As another measure of event size, seismic energy may provide some information related to the frictional property on the plate interface. We estimated radiated seismic energy for 620 repeating earthquakes of MJMA from 2.5 to 5.9, detected by the method of Uchida and Matsuzawa , in the Tohoku-Oki region. The study period is from 2001 to 2013, extending before and after the 2011 Tohoku-Oki earthquake of Mw 9, which is also accompanied with large afterslip [e.g., Ozawa et al., 2012]. The seismograms recorded by NIED Hi-net were used. We measured coda wave amplitude by the method of Mayeda et al.  and estimated source spectra and radiated seismic energy by the method of Baltay et al.  after slight modifications.
The estimated scaled energy, the ratio between radiated seismic energy and seismic moment, shows a slight increase with seismic moment. The scaled energy increases with depth, while its temporal change before and after the Tohoku-Oki earthquake is not systematic. The scaled energy also increases with the inter-event time of repeating earthquakes. This might be explained by the difference of fault strength, proportional to the logarithm of time. In addition to this healing relation, scaling relationship between seismic moment and the inter-event time of repeating earthquake is well known [Nadeau and Johnson, 1998]. From these healing and scaling relationships, it is expected that scaled energy is proportional to the logarithm of seismic moment. This prediction is generally consistent with our observation, though the moment dependency is too small to be recognized as power or log. This healing-related scaling may be applicable to general earthquakes, and might be associated with the unsolved problem of whether scaled energy is constant or size-dependent.