11:00 AM - 1:00 PM
[STT38-P06] Simple procedure of high frequency seismic signal duration measurement
Keywords:high frequency, seismic signal duration
Hara (2021, JpGU) presented a procedure to measure high frequency seismic signal duration (HFSD) in which the time when the amplitude of the smoothed band-pass filtered squared time series exceeded the threshold value (5 per cent) is used instead of picked P arrival times. He showed a good correlation between the measured durations and HFSDs obtained by the procedures of Hara (2007, EPS; 20019, SSJ). In the present study, we used theoretical P arrival times instead of picked P arrival times as another simpler attempt to measure HFSDs without picked P arrival times.
We used waveform data for large (Mw is equal to or greater than 7.2) shallow (focal depths are up to 50 km) events that occurred in between 1994 and May, 2021 at GSN stations in the epicentral distance range between 30 and 85 degrees. The number of the events is 152. The data were retrieved from the IRIS DMC.
The pre-processing of time series follows that of Hara (2007); the waveform data are band-pass (2-4 Hz) filtered, squared, and smoothed using the moving window average. Then, the measurement procedure consists of two steps. In the first step, duration is that from the theoretical P arrival time to the time when the amplitude of the smoothed times series become less than 25 % of the maximum value. Theoretical P arrival times are calculated using the TauP Toolkit (Crotwell et al., 1999) for iasp91 (Kennett and Engdahl, 1991). In the second step, the median of the HFSDs from the first step is used for smoothing time series following Hara (2019).
The measured HFSDs are more consistent with those using the picked P arrival times than those obtained by the procedure of Hara (2021). This suggests that the simple procedure of the present study could be used for automatic HFSD measurements. The durations obtained by Hara (2021) may be used to infer late energy release during rupture processes.
We used waveform data for large (Mw is equal to or greater than 7.2) shallow (focal depths are up to 50 km) events that occurred in between 1994 and May, 2021 at GSN stations in the epicentral distance range between 30 and 85 degrees. The number of the events is 152. The data were retrieved from the IRIS DMC.
The pre-processing of time series follows that of Hara (2007); the waveform data are band-pass (2-4 Hz) filtered, squared, and smoothed using the moving window average. Then, the measurement procedure consists of two steps. In the first step, duration is that from the theoretical P arrival time to the time when the amplitude of the smoothed times series become less than 25 % of the maximum value. Theoretical P arrival times are calculated using the TauP Toolkit (Crotwell et al., 1999) for iasp91 (Kennett and Engdahl, 1991). In the second step, the median of the HFSDs from the first step is used for smoothing time series following Hara (2019).
The measured HFSDs are more consistent with those using the picked P arrival times than those obtained by the procedure of Hara (2021). This suggests that the simple procedure of the present study could be used for automatic HFSD measurements. The durations obtained by Hara (2021) may be used to infer late energy release during rupture processes.