09:00 〜 09:15
[STT41-01] Duration measurements of synthetic WWSSN short-period seismograms and their comparison to centroid time shifts
キーワード：震源時間、合成WWSSN短周期地震波形、セントロイドタイムシフト
We measured durations of P-wave trains using synthetic World-Wide Standard Seismographic Network (WWSSN) short-period seismograms calculated from vertical-component broadband waveform data for large shallow earthquakes that occurred between 2000 and 2021. We selected 19 events with the centroid depths up to 50 km and the moment magnitudes greater than or equal to 8 in the Global Centroid-Moment-Tensor Catalog (Dziewonski et al. 1981; Ekström et al. 2012). We retrieved BHZ channel waveform data recorded at the Global Seismographic Network stations using the "FetchData" script (https://earthscope.github.io/fetch-scripts) using the earthquake information from the Advanced National Seismic System Comprehensive Earthquake Catalog (ComCat) (U.S. Geological Survey, Earthquake Hazards Program, 2017). We used data recorded at the stations in the epicentral distance range from 30 to 85 degrees.
We calculated synthetic WWSSN short-period seismograms through deconvolution using the instrumental responses and convolution of the WWSSN short-period seismograph frequency response which is from Bormann and Dewey (2014). We used the passband from 0.5 Hz to 5 Hz. We used the SAC (Goldstein et al. 2003; Goldstein and Snoke 2005) for these calculations. Following Houston and Kanamori (1986), we used 25 % as the threshold to measure durations. For each event, we used the mean of all the duration measurements for that event as the duration.
The duration for the December 26, 2004 Sumatra earthquake is 512.7 s. Since the duration for the1960 Chile obtained by Houston and Kanamori (1986) is 344.4 s, the duration of the 2004 Sumatra earthquake is longer than that of the1960 Chile when they are measured on the synthetic WWSSN short-period seismograms and observed short-period seismograms including WWSSN short-period seismograms, respectively.
We compared the measured durations to twice the centroid time shifts, which are representative of source times (Duputel et al., 2013). We used the centroid time shift data from the Global CMT Catalog, and the ComCat. The latter holds the centroid time shifts determined by the W phase inversions. We chose "the most preferred data available," which consisted of the data from Duputel et al. (2013) and those by the USGS. We excluded the 2004 Sumatra earthquake from this comparison, because the source duration of this event is exceptional with respect to the centroid time shift, as Duputel et al. (2013) indicated. The relationship obtained in this study is consistent with that obtained by Houston and Kanamori (1986). As a further attempt, we changed the threshold for duration measurements. We show the possibility of reducing the measurement uncertainties by choosing a higher threshold.
We calculated synthetic WWSSN short-period seismograms through deconvolution using the instrumental responses and convolution of the WWSSN short-period seismograph frequency response which is from Bormann and Dewey (2014). We used the passband from 0.5 Hz to 5 Hz. We used the SAC (Goldstein et al. 2003; Goldstein and Snoke 2005) for these calculations. Following Houston and Kanamori (1986), we used 25 % as the threshold to measure durations. For each event, we used the mean of all the duration measurements for that event as the duration.
The duration for the December 26, 2004 Sumatra earthquake is 512.7 s. Since the duration for the1960 Chile obtained by Houston and Kanamori (1986) is 344.4 s, the duration of the 2004 Sumatra earthquake is longer than that of the1960 Chile when they are measured on the synthetic WWSSN short-period seismograms and observed short-period seismograms including WWSSN short-period seismograms, respectively.
We compared the measured durations to twice the centroid time shifts, which are representative of source times (Duputel et al., 2013). We used the centroid time shift data from the Global CMT Catalog, and the ComCat. The latter holds the centroid time shifts determined by the W phase inversions. We chose "the most preferred data available," which consisted of the data from Duputel et al. (2013) and those by the USGS. We excluded the 2004 Sumatra earthquake from this comparison, because the source duration of this event is exceptional with respect to the centroid time shift, as Duputel et al. (2013) indicated. The relationship obtained in this study is consistent with that obtained by Houston and Kanamori (1986). As a further attempt, we changed the threshold for duration measurements. We show the possibility of reducing the measurement uncertainties by choosing a higher threshold.