9:15 AM - 9:30 AM
[SSS04-02] Rupture complexity of Mw3-7 earthquakes in Japan and its appearance in the time and frequency domains
★Invited Papers
Keywords:Source complexity, Source time function, Source spectrum, Earthquake scaling characteristics, Radiated energy
The omega-squared (ω2) source model and the corner frequency (fc) of the displacement spectrum have been widely used to estimate earthquake source parameters. The source spatial scale (a) is usually indirectly estimated from fc with the assumption that fc is inversely proportional to a. However, if the rupture pattern is complex, the fc vs. a relationship is non-unique. In this presentation, I will discuss the results and implications of our recent study (Yoshida and Kanamori, in prep.; SSJ, 2021) that systematically addresses the source complexity of 1700 Mw3-7 earthquakes by applying time-domain deconvolution methods to high-quality seismic network data from Japan.
We first determine the apparent moment-rate functions (AMRFs) of the earthquakes using the empirical Green's functions. Some AMRFs have multiple peaks, suggesting complex ruptures at multiple patches. The spectra of such complex events deviate significantly from the ω2 model, and fc is poorly defined. We use the AMRFs to estimate ER of the events having more than ten reliable AMRFs. We find no strong dependence of the scaled energy (eR=ER/M0) on the seismic moment (M0), focal mechanisms, or depth. The scaled energy varies by about one order of magnitude between earthquakes, with a median of about 3.5x10-5, which is comparable to the values obtained by previous studies.
We then measure the source complexity by the radiated energy enhancement factor (REEF; Ye et al., 2016) defined by the ratio of ER to the minimum scaled energy for the given M0 and source duration (T). REEF does not show strong scale dependence for Mw3-7 earthquakes. REEF varies approximately from 1 to 50, reflecting the diversity of rupture complexity. About a third of the events have REEF>5 and show multiple pulses on the time domain and complex shapes in the frequency domain. For such complex events, there is a significant deviation in the relationship between eR and M0/T3 from that expected from a simple source model (e.g., Brune, 1970).
Since there is no obvious way to relate T to the spatial scale for complex events, stress drop and radiation efficiency are indeterminate. For Brune's (1970) model, the radiation efficiency is 0.47 regardless of stress drop. Despite the ambiguity of stress drop, the radiation efficiency of 0.47 appears to be a reasonable rough average for ordinary brittle events in the crust. If we assume that simple and complex events have the same radiation efficiency, on average, in the crust of similar tectonic environments, a rough estimate of stress drop is given by eR.
We first determine the apparent moment-rate functions (AMRFs) of the earthquakes using the empirical Green's functions. Some AMRFs have multiple peaks, suggesting complex ruptures at multiple patches. The spectra of such complex events deviate significantly from the ω2 model, and fc is poorly defined. We use the AMRFs to estimate ER of the events having more than ten reliable AMRFs. We find no strong dependence of the scaled energy (eR=ER/M0) on the seismic moment (M0), focal mechanisms, or depth. The scaled energy varies by about one order of magnitude between earthquakes, with a median of about 3.5x10-5, which is comparable to the values obtained by previous studies.
We then measure the source complexity by the radiated energy enhancement factor (REEF; Ye et al., 2016) defined by the ratio of ER to the minimum scaled energy for the given M0 and source duration (T). REEF does not show strong scale dependence for Mw3-7 earthquakes. REEF varies approximately from 1 to 50, reflecting the diversity of rupture complexity. About a third of the events have REEF>5 and show multiple pulses on the time domain and complex shapes in the frequency domain. For such complex events, there is a significant deviation in the relationship between eR and M0/T3 from that expected from a simple source model (e.g., Brune, 1970).
Since there is no obvious way to relate T to the spatial scale for complex events, stress drop and radiation efficiency are indeterminate. For Brune's (1970) model, the radiation efficiency is 0.47 regardless of stress drop. Despite the ambiguity of stress drop, the radiation efficiency of 0.47 appears to be a reasonable rough average for ordinary brittle events in the crust. If we assume that simple and complex events have the same radiation efficiency, on average, in the crust of similar tectonic environments, a rough estimate of stress drop is given by eR.